shinuwa/codex-reset-alerts
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Initial commit

Browse source
This commit is contained in:
Shinuwa 2026-08-17 12:12:01 +02:00
commit 2c78bb9c76
201 changed files with 196806 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

1
.codex-home/skills/.system/.codex-system-skills.marker Normal file
View file

@ -0,0 +1 @@
7e337e3e918d2aa5

201
.codex-home/skills/.system/imagegen/LICENSE.txt Normal file
View file

@ -0,0 +1,201 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf of
any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don\'t include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

315
.codex-home/skills/.system/imagegen/SKILL.md Normal file
View file

@ -0,0 +1,315 @@
---
name: "imagegen"
description: "Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output should be a bitmap asset rather than repo-native code or vector. Do not use when the task is better handled by editing existing SVG/vector/code-native assets, extending an established icon or logo system, or building the visual directly in HTML/CSS/canvas."
---
# Image Generation Skill
Generates or edits images for the current project (for example website assets, game assets, UI mockups, product mockups, wireframes, logo design, photorealistic images, or infographics).
## Top-level modes and rules
This skill has exactly two top-level modes:
- **Default built-in tool mode (preferred):** built-in `image_gen` tool for image generation, editing, and transparent-image requests. Does not require `OPENAI_API_KEY`.
- **Fallback CLI mode:** `scripts/image_gen.py` CLI. Use when the user explicitly asks for or confirms the CLI/API/model path. Requires `OPENAI_API_KEY`.
Within CLI fallback, the CLI exposes three subcommands:
- `generate`
- `edit`
- `generate-batch`
Rules:
- Use the built-in `image_gen` tool by default for normal image generation and editing requests.
- Do not switch to CLI fallback for ordinary quality, size, or file-path control.
- For transparent images, ask built-in `image_gen` for a transparent background and preserve the generated alpha.
- Never silently switch from built-in `image_gen` or CLI `gpt-image-2` to CLI `gpt-image-1.5`; ask the user first unless they explicitly requested `gpt-image-1.5`.
- The word `batch` by itself does not mean CLI fallback. If the user asks for many assets or says to batch-generate assets without explicitly asking for CLI/API/model controls, stay on the built-in path and issue one built-in call per requested asset or variant.
- If the built-in tool fails or is unavailable, tell the user the CLI fallback exists and that it requires `OPENAI_API_KEY`. Proceed only if the user explicitly asks for that fallback.
- If the user explicitly asks for CLI mode, use the bundled `scripts/image_gen.py` workflow. Do not create one-off SDK runners.
- Never modify `scripts/image_gen.py`. If something is missing, ask the user before doing anything else.
Built-in save-path policy:
- In built-in tool mode, Codex saves generated images under `$CODEX_HOME/*` by default.
- Do not describe or rely on OS temp as the default built-in destination.
- Do not describe or rely on a destination-path argument (if any) on the built-in `image_gen` tool. If a specific location is needed, generate first and then move or copy the selected output from `$CODEX_HOME/generated_images/...`.
- Save-path precedence in built-in mode:
1. If the user names a destination, move or copy the selected output there.
2. If the image is meant for the current project, move or copy the final selected image into the workspace before finishing.
3. If the image is only for preview or brainstorming, render it inline; the underlying file can remain at the default `$CODEX_HOME/*` path.
- Never leave a project-referenced asset only at the default `$CODEX_HOME/*` path.
- Do not overwrite an existing asset unless the user explicitly asked for replacement; otherwise create a sibling versioned filename such as `hero-v2.png` or `item-icon-edited.png`.
Shared prompt guidance for both modes lives in `references/prompting.md` and `references/sample-prompts.md`.
Fallback-only docs/resources for CLI mode:
- `references/cli.md`
- `references/image-api.md`
- `references/codex-network.md`
- `scripts/image_gen.py`
## When to use
- Generate a new image (concept art, product shot, cover, website hero)
- Generate a new image using one or more reference images for style, composition, or mood
- Edit an existing image (inpainting, lighting or weather transformations, background replacement, object removal, compositing, transparent background)
- Produce many assets or variants for one task
## When not to use
- Extending or matching an existing SVG/vector icon set, logo system, or illustration library inside the repo
- Creating simple shapes, diagrams, wireframes, or icons that are better produced directly in SVG, HTML/CSS, or canvas
- Making a small project-local asset edit when the source file already exists in an editable native format
- Any task where the user clearly wants deterministic code-native output instead of a generated bitmap
## Decision tree
Think about two separate questions:
1. **Intent:** is this a new image or an edit of an existing image?
2. **Execution strategy:** is this one asset or many assets/variants?
Intent:
- If the user wants to modify an existing image while preserving parts of it, treat the request as **edit**.
- If the user provides images only as references for style, composition, mood, or subject guidance, treat the request as **generate**.
- If the user provides no images, treat the request as **generate**.
Built-in edit semantics:
- Built-in edit mode is for images already visible in the conversation context, such as attached images or images generated earlier in the thread.
- If the user wants to edit a local image file with the built-in tool, first load it with built-in `view_image` tool so the image is visible in the conversation context, then proceed with the built-in edit flow.
- Do not promise arbitrary filesystem-path editing through the built-in tool.
- If a local file still needs direct file-path control, masks, or other explicit CLI-only parameters, use the explicit CLI fallback only when the user asks for it.
- For edits, preserve invariants aggressively and save non-destructively by default.
Execution strategy:
- In the built-in default path, produce many assets or variants by issuing one `image_gen` call per requested asset or variant.
- In the CLI fallback path, use the CLI `generate-batch` subcommand only when the user explicitly chose CLI mode and needs many prompts/assets.
- For many distinct assets, do not use `n` as a substitute for separate prompts. `n` is for variants of one prompt; distinct assets need distinct built-in calls or distinct CLI `generate-batch` jobs.
Assume the user wants a new image unless they clearly ask to change an existing one.
## Workflow
1. Decide the top-level mode: built-in by default, including transparent-output requests; fallback CLI only if explicitly requested or confirmed.
2. Decide the intent: `generate` or `edit`.
3. Decide whether the output is preview-only or meant to be consumed by the current project.
4. Decide the execution strategy: single asset vs repeated built-in calls vs CLI `generate-batch`.
5. Collect inputs up front: prompt(s), exact text (verbatim), constraints/avoid list, and any input images.
6. For every input image, label its role explicitly:
- reference image
- edit target
- supporting insert/style/compositing input
7. If the edit target is only on the local filesystem and you are staying on the built-in path, inspect it with `view_image` first so the image is available in conversation context.
8. If the user asked for a photo, illustration, sprite, product image, banner, or other explicitly raster-style asset, use `image_gen` rather than substituting SVG/HTML/CSS placeholders. If the request is for an icon, logo, or UI graphic that should match existing repo-native SVG/vector/code assets, prefer editing those directly instead.
9. Augment the prompt based on specificity:
- If the user's prompt is already specific and detailed, normalize it into a clear spec without adding creative requirements.
- If the user's prompt is generic, add tasteful augmentation only when it materially improves output quality.
10. Use the built-in `image_gen` tool by default.
11. For transparent-output requests, ask built-in `image_gen` for a transparent background and preserve the generated alpha channel.
12. Inspect outputs and validate: subject, style, composition, text accuracy, and invariants/avoid items.
13. Iterate with a single targeted change, then re-check.
14. For preview-only work, render the image inline; the underlying file may remain at the default `$CODEX_HOME/generated_images/...` path.
15. For project-bound work, move or copy the selected artifact into the workspace and update any consuming code or references. Never leave a project-referenced asset only at the default `$CODEX_HOME/generated_images/...` path.
16. For batches or multi-asset requests, persist every requested deliverable final in the workspace unless the user explicitly asked to keep outputs preview-only. Discarded variants do not need to be kept unless requested.
17. If the user explicitly chooses or confirms the CLI fallback, then use the fallback-only docs for model, quality, size, `input_fidelity`, masks, output format, output paths, and network setup.
18. Always report the final saved path(s) for any workspace-bound asset(s), plus the final prompt or prompt set and whether the built-in tool or fallback CLI mode was used.
## Transparent image requests
Ask built-in `image_gen` for a genuinely transparent background and preserve its alpha.
## Prompt augmentation
Reformat user prompts into a structured, production-oriented spec. Make the user's goal clearer and more actionable, but do not blindly add detail.
Treat this as prompt-shaping guidance, not a closed schema. Use only the lines that help, and add a short extra labeled line when it materially improves clarity.
### Specificity policy
Use the user's prompt specificity to decide how much augmentation is appropriate:
- If the prompt is already specific and detailed, preserve that specificity and only normalize/structure it.
- If the prompt is generic, you may add tasteful augmentation when it will materially improve the result.
Allowed augmentations:
- composition or framing hints
- polish level or intended-use hints
- practical layout guidance
- reasonable scene concreteness that supports the stated request
Not allowed augmentations:
- extra characters or objects that are not implied by the request
- brand names, slogans, palettes, or narrative beats that are not implied
- arbitrary side-specific placement unless the surrounding layout supports it
## Use-case taxonomy (exact slugs)
Classify each request into one of these buckets and keep the slug consistent across prompts and references.
Generate:
- photorealistic-natural — candid/editorial lifestyle scenes with real texture and natural lighting.
- product-mockup — product/packaging shots, catalog imagery, merch concepts.
- ui-mockup — app/web interface mockups and wireframes; specify the desired fidelity.
- infographic-diagram — diagrams/infographics with structured layout and text.
- scientific-educational — classroom explainers, scientific diagrams, and learning visuals with required labels and accuracy constraints.
- ads-marketing — campaign concepts and ad creatives with audience, brand position, scene, and exact tagline/copy.
- productivity-visual — slide, chart, workflow, and data-heavy business visuals.
- logo-brand — logo/mark exploration, vector-friendly.
- illustration-story — comics, children’s book art, narrative scenes.
- stylized-concept — style-driven concept art, 3D/stylized renders.
- historical-scene — period-accurate/world-knowledge scenes.
Edit:
- text-localization — translate/replace in-image text, preserve layout.
- identity-preserve — try-on, person-in-scene; lock face/body/pose.
- precise-object-edit — remove/replace a specific element (including interior swaps).
- lighting-weather — time-of-day/season/atmosphere changes only.
- background-extraction — transparent background / clean cutout. Ask built-in `image_gen` for actual transparency.
- style-transfer — apply reference style while changing subject/scene.
- compositing — multi-image insert/merge with matched lighting/perspective.
- sketch-to-render — drawing/line art to photoreal render.
## Shared prompt schema
Use the following labeled spec as shared prompt scaffolding for both top-level modes:
```text
Use case: <taxonomy slug>
Asset type: <where the asset will be used>
Primary request: <user's main prompt>
Input images: <Image 1: role; Image 2: role> (optional)
Scene/backdrop: <environment>
Subject: <main subject>
Style/medium: <photo/illustration/3D/etc>
Composition/framing: <wide/close/top-down; placement>
Lighting/mood: <lighting + mood>
Color palette: <palette notes>
Materials/textures: <surface details>
Text (verbatim): "<exact text>"
Constraints: <must keep/must avoid>
Avoid: <negative constraints>
```
Notes:
- `Asset type` and `Input images` are prompt scaffolding, not dedicated CLI flags.
- `Scene/backdrop` refers to the visual setting. It is not the same as the fallback CLI `background` parameter, which controls output transparency behavior.
- Fallback-only execution notes such as `Quality:`, `Input fidelity:`, masks, output format, and output paths belong in the CLI path only. Do not treat them as built-in `image_gen` tool arguments.
Augmentation rules:
- Keep it short.
- Add only the details needed to improve the prompt materially.
- For edits, explicitly list invariants (`change only X; keep Y unchanged`).
- If any critical detail is missing and blocks success, ask a question; otherwise proceed.
## Examples
### Generation example (hero image)
```text
Use case: product-mockup
Asset type: landing page hero
Primary request: a minimal hero image of a ceramic coffee mug
Style/medium: clean product photography
Composition/framing: wide composition with usable negative space for page copy if needed
Lighting/mood: soft studio lighting
Constraints: no logos, no text, no watermark
```
### Edit example (invariants)
```text
Use case: precise-object-edit
Asset type: product photo background replacement
Primary request: replace only the background with a warm sunset gradient
Constraints: change only the background; keep the product and its edges unchanged; no text; no watermark
```
## Prompting best practices
- Structure prompt as scene/backdrop -> subject -> details -> constraints.
- Include intended use (ad, UI mock, infographic) to set the mode and polish level.
- Use camera/composition language for photorealism.
- Only use SVG/vector stand-ins when the user explicitly asked for vector output or a non-image placeholder.
- Quote exact text and specify typography + placement.
- For tricky words, spell them letter-by-letter and require verbatim rendering.
- For multi-image inputs, reference images by index and describe how they should be used.
- For edits, repeat invariants every iteration to reduce drift.
- Iterate with single-change follow-ups.
- If the prompt is generic, add only the extra detail that will materially help.
- If the prompt is already detailed, normalize it instead of expanding it.
- For CLI fallback only, see `references/cli.md` and `references/image-api.md` for model, `quality`, `input_fidelity`, masks, output format, and output-path guidance.
- For transparent images, ask built-in `image_gen` for actual transparency and preserve its alpha.
More principles shared by both modes: `references/prompting.md`.
Copy/paste specs shared by both modes: `references/sample-prompts.md`.
## Guidance by asset type
Asset-type templates (website assets, game assets, wireframes, logo) are consolidated in `references/sample-prompts.md`.
## gpt-image-2 guidance for CLI fallback
The fallback CLI defaults to `gpt-image-2`.
- Use `gpt-image-2` for new CLI/API workflows unless the user confirms a different model.
- CLI `gpt-image-2` does not support `background=transparent`; ask before using `gpt-image-1.5` unless the user explicitly requested that model.
- `gpt-image-2` always uses high fidelity for image inputs; do not set `input_fidelity` with this model.
- `gpt-image-2` supports `quality` values `low`, `medium`, `high`, and `auto`.
- Use `quality low` for fast drafts, thumbnails, and quick iterations. Use `medium`, `high`, or `auto` for final assets, dense text, diagrams, identity-sensitive edits, or high-resolution outputs.
- Square images are typically fastest to generate. Use `1024x1024` for fast square drafts.
- If the user asks for 4K-style output, use `3840x2160` for landscape or `2160x3840` for portrait.
- `gpt-image-2` size may be `auto` or `WIDTHxHEIGHT` if all constraints hold: max edge `<= 3840px`, both edges multiples of `16px`, long-to-short ratio `<= 3:1`, total pixels between `655,360` and `8,294,400`.
Popular `gpt-image-2` sizes:
- `1024x1024` square
- `1536x1024` landscape
- `1024x1536` portrait
- `2048x2048` 2K square
- `2048x1152` 2K landscape
- `3840x2160` 4K landscape
- `2160x3840` 4K portrait
- `auto`
## Fallback CLI mode only
### Temp and output conventions
These conventions apply only to the CLI fallback. They do not describe built-in `image_gen` output behavior.
- Use `tmp/imagegen/` for intermediate files (for example JSONL batches); delete them when done.
- Write final artifacts under `output/imagegen/`.
- Use `--out` or `--out-dir` to control output paths; keep filenames stable and descriptive.
### Dependencies
Prefer `uv` for dependency management in this repo.
Required Python package:
```bash
uv pip install openai
```
Optional for image inspection and downscaling:
```bash
uv pip install pillow
```
Portability note:
- If you are using the installed skill outside this repo, install dependencies into that environment with its package manager.
- In uv-managed environments, `uv pip install ...` remains the preferred path.
### Environment
- `OPENAI_API_KEY` must be set for live API calls.
- Do not ask the user for `OPENAI_API_KEY` when using the built-in `image_gen` tool.
- Never ask the user to paste the full key in chat. Ask them to set it locally and confirm when ready.
If the key is missing, give the user these steps:
1. Create an API key in the OpenAI platform UI: https://platform.openai.com/api-keys
2. Set `OPENAI_API_KEY` as an environment variable in their system.
3. Offer to guide them through setting the environment variable for their OS/shell if needed.
If installation is not possible in this environment, tell the user which dependency is missing and how to install it into their active environment.
### Script-mode notes
- CLI commands + examples: `references/cli.md`
- API parameter quick reference: `references/image-api.md`
- Network approvals / sandbox settings for CLI mode: `references/codex-network.md`
## Reference map
- `references/prompting.md`: shared prompting principles for both modes.
- `references/sample-prompts.md`: shared copy/paste prompt recipes for both modes.
- `references/cli.md`: fallback-only CLI usage via `scripts/image_gen.py`.
- `references/image-api.md`: fallback-only API/CLI parameter reference.
- `references/codex-network.md`: fallback-only network/sandbox troubleshooting for CLI mode.
- `scripts/image_gen.py`: fallback-only CLI implementation. Use only when the user explicitly chooses or confirms CLI mode.

6
.codex-home/skills/.system/imagegen/agents/openai.yaml Normal file
View file

@ -0,0 +1,6 @@
interface:
display_name: "Image Gen"
short_description: "Generate or edit images for websites, games, and more"
icon_small: "./assets/imagegen-small.svg"
icon_large: "./assets/imagegen.png"
default_prompt: "Use $imagegen to make or edit an image for this project."

5
.codex-home/skills/.system/imagegen/assets/imagegen-small.svg Normal file
View file

@ -0,0 +1,5 @@
<svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" fill="currentColor" viewBox="0 0 16 16">
<path fill="currentColor" d="M7.51 6.827a1 1 0 1 1 .278 1.982 1 1 0 0 1-.278-1.982Z"/>
<path fill="currentColor" fill-rule="evenodd" d="M8.31 4.47c.368-.016.699.008 1.016.124l.186.075c.423.194.786.5 1.047.888l.067.107c.148.253.235.533.3.848.073.354.126.797.193 1.343l.277 2.25.088.745c.024.224.041.425.049.605.013.322-.004.615-.085.896l-.04.12a2.53 2.53 0 0 1-.802 1.115l-.16.118c-.281.189-.596.292-.956.366a9.46 9.46 0 0 1-.6.1l-.743.094-2.25.277c-.547.067-.99.121-1.35.136a2.765 2.765 0 0 1-.896-.085l-.12-.039a2.533 2.533 0 0 1-1.115-.802l-.118-.161c-.189-.28-.292-.596-.366-.956a9.42 9.42 0 0 1-.1-.599l-.094-.744-.276-2.25a17.884 17.884 0 0 1-.137-1.35c-.015-.367.009-.698.124-1.015l.076-.185c.193-.423.5-.787.887-1.048l.107-.067c.253-.148.534-.234.849-.3.354-.073.796-.126 1.343-.193l2.25-.277.744-.088c.224-.024.425-.041.606-.049Zm-2.905 5.978a1.47 1.47 0 0 0-.875.074c-.127.052-.267.146-.475.344-.212.204-.462.484-.822.889l-.314.351c.018.115.036.219.055.313.061.295.127.458.206.575l.07.094c.167.211.39.372.645.465l.109.032c.119.027.273.038.499.029.308-.013.7-.06 1.264-.13l2.25-.275.727-.093.198-.03-2.05-1.64a16.848 16.848 0 0 0-.96-.738c-.18-.121-.31-.19-.421-.23l-.106-.03Zm2.95-4.915c-.154.006-.33.021-.536.043l-.729.086-2.25.276c-.564.07-.956.118-1.257.18a1.937 1.937 0 0 0-.478.15l-.097.057a1.47 1.47 0 0 0-.515.608l-.044.107c-.048.133-.073.307-.06.608.012.307.06.7.129 1.264l.22 1.8.178-.197c.145-.159.278-.298.403-.418.255-.243.507-.437.809-.56l.181-.067a2.526 2.526 0 0 1 1.328-.06l.118.029c.27.079.517.215.772.387.287.194.619.46 1.03.789l2.52 2.016c.146-.148.26-.326.332-.524l.031-.109c.027-.119.039-.273.03-.499a8.311 8.311 0 0 0-.044-.536l-.086-.728-.276-2.25c-.07-.564-.118-.956-.18-1.258a1.935 1.935 0 0 0-.15-.477l-.057-.098a1.468 1.468 0 0 0-.608-.515l-.107-.043c-.133-.049-.306-.074-.607-.061Z" clip-rule="evenodd"/>
<path fill="currentColor" d="M7.783 1.272c.36.014.803.07 1.35.136l2.25.277.743.095c.224.03.423.062.6.099.36.074.675.177.955.366l.161.118c.364.29.642.675.802 1.115l.04.12c.081.28.098.574.085.896a9.42 9.42 0 0 1-.05.605l-.087.745-.277 2.25c-.067.547-.12.989-.193 1.343a2.765 2.765 0 0 1-.3.848l-.067.107a2.534 2.534 0 0 1-.415.474l-.086.064a.532.532 0 0 1-.622-.858l.13-.13c.04-.046.077-.094.111-.145l.057-.098c.055-.109.104-.256.15-.477.062-.302.11-.694.18-1.258l.276-2.25.086-.728c.022-.207.037-.382.043-.536.01-.226-.002-.38-.029-.5l-.032-.108a1.469 1.469 0 0 0-.464-.646l-.094-.069c-.118-.08-.28-.145-.575-.206a8.285 8.285 0 0 0-.53-.088l-.728-.092-2.25-.276c-.565-.07-.956-.117-1.264-.13a1.94 1.94 0 0 0-.5.029l-.108.032a1.469 1.469 0 0 0-.647.465l-.068.094c-.054.08-.102.18-.146.33l-.04.1a.533.533 0 0 1-.98-.403l.055-.166c.059-.162.133-.314.23-.457l.117-.16c.29-.365.675-.643 1.115-.803l.12-.04c.28-.08.574-.097.896-.084Z"/>
</svg>
Side by side

After

Width:  |  Height:  |  Size: 2.8 KiB

BIN
.codex-home/skills/.system/imagegen/assets/imagegen.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 1.7 KiB

242
.codex-home/skills/.system/imagegen/references/cli.md Normal file
View file

@ -0,0 +1,242 @@
# CLI reference (`scripts/image_gen.py`)
This file is for the fallback CLI mode only. Read it when the user explicitly asks to use `scripts/image_gen.py` / CLI / API / model controls, or after the user explicitly confirms that a transparent-output request should use the `gpt-image-1.5` true-transparency fallback path.
`generate-batch` is a CLI subcommand in this fallback path. It is not a top-level mode of the skill.
The word `batch` in a user request is not CLI opt-in by itself.
## What this CLI does
- `generate`: generate a new image from a prompt
- `edit`: edit one or more existing images
- `generate-batch`: run many generation jobs from a JSONL file after the user explicitly chooses CLI/API/model controls
Real API calls require **network access** + `OPENAI_API_KEY`. `--dry-run` does not.
## Quick start (works from any repo)
Set a stable path to the skill CLI (default `CODEX_HOME` is `~/.codex`):
```
export CODEX_HOME="${CODEX_HOME:-$HOME/.codex}"
export IMAGE_GEN="$CODEX_HOME/skills/.system/imagegen/scripts/image_gen.py"
```
Install dependencies into that environment with its package manager. In uv-managed environments, `uv pip install ...` remains the preferred path.
## Quick start
Dry-run (no API call; no network required; does not require the `openai` package):
```bash
python "$IMAGE_GEN" generate \
--prompt "Test" \
--out output/imagegen/test.png \
--dry-run
```
Notes:
- One-off dry-runs print the API payload and the computed output path(s).
- Repo-local finals should live under `output/imagegen/`.
Generate (requires `OPENAI_API_KEY` + network):
```bash
python "$IMAGE_GEN" generate \
--prompt "A cozy alpine cabin at dawn" \
--size 1024x1024 \
--out output/imagegen/alpine-cabin.png
```
Edit:
```bash
python "$IMAGE_GEN" edit \
--image input.png \
--prompt "Replace only the background with a warm sunset" \
--out output/imagegen/sunset-edit.png
```
## Guardrails
- Use the bundled CLI directly (`python "$IMAGE_GEN" ...`) after activating the correct environment.
- Do **not** create one-off runners (for example `gen_images.py`) unless the user explicitly asks for a custom wrapper.
- **Never modify** `scripts/image_gen.py`. If something is missing, ask the user before doing anything else.
- Do not silently downgrade from CLI `gpt-image-2` or built-in `image_gen` to CLI `gpt-image-1.5`; ask first unless the user explicitly requested `gpt-image-1.5`.
## Defaults
- Model: `gpt-image-2`
- Supported model family for this CLI: GPT Image models (`gpt-image-*`)
- Size: `auto`
- Quality: `medium`
- Output format: `png`
- Default one-off output path: `output/imagegen/output.png`
- Background: unspecified unless `--background` is set
## gpt-image-2 size and model guidance
`gpt-image-2` is the default model for new CLI fallback work.
- Use `--quality low` for fast drafts, thumbnails, and quick iterations.
- Use `--quality medium`, `--quality high`, or `--quality auto` for final assets, dense text, diagrams, identity-sensitive edits, and high-resolution outputs.
- Square images are typically fastest. Use `--size 1024x1024` for quick square drafts.
- If the user asks for 4K-style output, use `--size 3840x2160` for landscape or `--size 2160x3840` for portrait.
- Do not pass `--input-fidelity` with `gpt-image-2`; this model always uses high fidelity for image inputs.
- Do not use `--background transparent` with CLI `gpt-image-2`; ask before using `gpt-image-1.5` unless the user explicitly requested that model.
Popular `gpt-image-2` sizes:
- `1024x1024`
- `1536x1024`
- `1024x1536`
- `2048x2048`
- `2048x1152`
- `3840x2160`
- `2160x3840`
- `auto`
`gpt-image-2` size constraints:
- max edge `<= 3840px`
- both edges multiples of `16px`
- long edge to short edge ratio `<= 3:1`
- total pixels between `655,360` and `8,294,400`
- outputs above `2560x1440` total pixels are experimental
Fast draft:
```bash
python "$IMAGE_GEN" generate \
--prompt "A product thumbnail of a matte ceramic mug on a stone surface" \
--quality low \
--size 1024x1024 \
--out output/imagegen/mug-draft.png
```
Final 2K landscape:
```bash
python "$IMAGE_GEN" generate \
--prompt "A polished landing-page hero image of a matte ceramic mug on a stone surface" \
--quality high \
--size 2048x1152 \
--out output/imagegen/mug-hero.png
```
4K landscape:
```bash
python "$IMAGE_GEN" generate \
--prompt "A detailed architectural visualization at golden hour" \
--size 3840x2160 \
--quality high \
--out output/imagegen/architecture-4k.png
```
True transparent fallback request:
Ask for confirmation before using this command unless the user explicitly requested `gpt-image-1.5`.
```bash
python "$IMAGE_GEN" generate \
--model gpt-image-1.5 \
--prompt "A clean product cutout on a transparent background" \
--background transparent \
--output-format png \
--out output/imagegen/product-cutout.png
```
Explain that CLI `gpt-image-2` does not support `background=transparent`, so transparent CLI output requires the confirmed `gpt-image-1.5` fallback.
## Quality, input fidelity, and masks (CLI fallback only)
These are explicit CLI controls. They are not built-in `image_gen` tool arguments.
- `--quality` works for `generate`, `edit`, and `generate-batch`: `low|medium|high|auto`
- `--input-fidelity` is **edit-only** and validated as `low|high`; it is not supported for `gpt-image-2`
- `--mask` is **edit-only**
Example:
```bash
python "$IMAGE_GEN" edit \
--model gpt-image-1.5 \
--image input.png \
--prompt "Change only the background" \
--quality high \
--input-fidelity high \
--out output/imagegen/background-edit.png
```
Mask notes:
- For multi-image edits, pass repeated `--image` flags. Their order is meaningful, so describe each image by index and role in the prompt.
- The CLI accepts a single `--mask`.
- Image and mask must be the same size and format and each under 50MB.
- Masks must include an alpha channel.
- If multiple input images are provided, the mask applies to the first image.
- Masking is prompt-guided; do not promise exact pixel-perfect mask boundaries.
- Use a PNG mask when possible; the script treats mask handling as best-effort and does not perform full preflight validation beyond file checks/warnings.
- In the edit prompt, repeat invariants (`change only the background; keep the subject unchanged`) to reduce drift.
## Output handling
- Use `tmp/imagegen/` for temporary JSONL inputs or scratch files.
- Use `output/imagegen/` for final outputs.
- Reruns fail if a target file already exists unless you pass `--force`.
- `--out-dir` changes one-off naming to `image_1.<ext>`, `image_2.<ext>`, and so on.
- Downscaled copies use the default suffix `-web` unless you override it.
## Common recipes
Generate with augmentation fields:
```bash
python "$IMAGE_GEN" generate \
--prompt "A minimal hero image of a ceramic coffee mug" \
--use-case "product-mockup" \
--style "clean product photography" \
--composition "wide product shot with usable negative space for page copy" \
--constraints "no logos, no text" \
--out output/imagegen/mug-hero.png
```
Generate + also write a downscaled copy for fast web loading:
```bash
python "$IMAGE_GEN" generate \
--prompt "A cozy alpine cabin at dawn" \
--size 1024x1024 \
--downscale-max-dim 1024 \
--out output/imagegen/alpine-cabin.png
```
Generate multiple prompts concurrently (async batch):
```bash
mkdir -p tmp/imagegen output/imagegen/batch
cat > tmp/imagegen/prompts.jsonl << 'EOF'
{"prompt":"Cavernous hangar interior with a compact shuttle parked near the center","use_case":"stylized-concept","composition":"wide-angle, low-angle","lighting":"volumetric light rays through drifting fog","constraints":"no logos or trademarks; no watermark","size":"1536x1024"}
{"prompt":"Gray wolf in profile in a snowy forest","use_case":"photorealistic-natural","composition":"eye-level","constraints":"no logos or trademarks; no watermark","size":"1024x1024"}
EOF
python "$IMAGE_GEN" generate-batch \
--input tmp/imagegen/prompts.jsonl \
--out-dir output/imagegen/batch \
--concurrency 5
rm -f tmp/imagegen/prompts.jsonl
```
Notes:
- `generate-batch` requires `--out-dir`.
- generate-batch requires --out-dir.
- Use `--concurrency` to control parallelism (default `5`).
- Per-job overrides are supported in JSONL (for example `size`, `quality`, `background`, `output_format`, `output_compression`, `moderation`, `n`, `model`, `out`, and prompt-augmentation fields).
- `--n` generates multiple variants for a single prompt; `generate-batch` is for many different prompts.
- In batch mode, per-job `out` is treated as a filename under `--out-dir`.
- For many requested deliverable assets, provide one prompt/job per distinct asset and use semantic filenames when possible.
## CLI notes
- Supported sizes depend on the model. `gpt-image-2` supports flexible constrained sizes; older GPT Image models support `1024x1024`, `1536x1024`, `1024x1536`, or `auto`.
- True transparent CLI outputs require `output_format` to be `png` or `webp` and are not supported by `gpt-image-2`.
- `--prompt-file`, `--output-compression`, `--moderation`, `--max-attempts`, `--fail-fast`, `--force`, and `--no-augment` are supported.
- This CLI is intended for GPT Image models. Do not assume older non-GPT image-model behavior applies here.
## See also
- API parameter quick reference for fallback CLI mode: `references/image-api.md`
- Prompt examples shared across both top-level modes: `references/sample-prompts.md`
- Network/sandbox notes for fallback CLI mode: `references/codex-network.md`
- Built-in-first transparent image workflow: `SKILL.md`

33
.codex-home/skills/.system/imagegen/references/codex-network.md Normal file
View file

@ -0,0 +1,33 @@
# Codex network approvals / sandbox notes
This file is for the fallback CLI mode only. Read it when the user explicitly asks to use `scripts/image_gen.py` / CLI / API / model controls, or after the user explicitly confirms that a transparent-output request should use the `gpt-image-1.5` true-transparency fallback path.
This guidance is intentionally isolated from `SKILL.md` because it can vary by environment and may become stale. Prefer the defaults in your environment when in doubt.
## Why am I asked to approve image generation calls?
The fallback CLI uses the OpenAI Image API, so it needs outbound network access. In many Codex setups, network access is disabled by default and/or the approval policy requires confirmation before networked commands run.
## Important note about approvals vs network
- `--ask-for-approval never` suppresses approval prompts.
- It does **not** by itself enable network access.
- In `workspace-write`, network access still depends on your Codex configuration (for example `[sandbox_workspace_write] network_access = true`).
## How do I reduce repeated approval prompts?
If you trust the repo and want fewer prompts, use a configuration or profile that both:
- enables network for the sandbox mode you plan to use
- sets an approval policy that matches your risk tolerance
Example `~/.codex/config.toml` pattern:
```toml
approval_policy = "on-request"
sandbox_mode = "workspace-write"
[sandbox_workspace_write]
network_access = true
```
If you want quieter automation after network is enabled, you can choose a stricter approval policy, but do that intentionally and with care.
## Safety note
Enabling network and reducing approvals lowers friction, but increases risk if you run untrusted code or work in an untrusted repository.

90
.codex-home/skills/.system/imagegen/references/image-api.md Normal file
View file

@ -0,0 +1,90 @@
# Image API quick reference
This file is for the fallback CLI mode only. Use it when the user explicitly asks to use `scripts/image_gen.py` / CLI / API / model controls, or after the user explicitly confirms that a transparent-output request should use the `gpt-image-1.5` true-transparency fallback path.
These parameters describe the Image API and bundled CLI fallback surface. Do not assume they are normal arguments on the built-in `image_gen` tool.
## Scope
- This fallback CLI is intended for GPT Image models (`gpt-image-2`, `gpt-image-1.5`, `gpt-image-1`, and `gpt-image-1-mini`).
- The built-in `image_gen` tool and the fallback CLI do not expose the same controls.
## Model summary
| Model | Quality | Input fidelity | Resolutions | Recommended use |
| --- | --- | --- | --- | --- |
| `gpt-image-2` | `low`, `medium`, `high`, `auto` | Always high fidelity for image inputs; do not set `input_fidelity` | `auto` or flexible sizes that satisfy the constraints below | Default for new CLI/API workflows: high-quality generation and editing, text-heavy images, photorealism, compositing, identity-sensitive edits, and workflows where fewer retries matter |
| `gpt-image-1.5` | `low`, `medium`, `high`, `auto` | `low`, `high` | `1024x1024`, `1024x1536`, `1536x1024`, `auto` | True transparent-background fallback and backward-compatible workflows |
| `gpt-image-1` | `low`, `medium`, `high`, `auto` | `low`, `high` | `1024x1024`, `1024x1536`, `1536x1024`, `auto` | Legacy compatibility |
| `gpt-image-1-mini` | `low`, `medium`, `high`, `auto` | `low`, `high` | `1024x1024`, `1024x1536`, `1536x1024`, `auto` | Cost-sensitive draft batches and lower-stakes previews |
## gpt-image-2 sizes
`gpt-image-2` accepts `auto` or any `WIDTHxHEIGHT` size that satisfies all constraints:
- Maximum edge length must be less than or equal to `3840px`.
- Both edges must be multiples of `16px`.
- Long edge to short edge ratio must not exceed `3:1`.
- Total pixels must be at least `655,360` and no more than `8,294,400`.
Popular sizes:
| Label | Size | Notes |
| --- | --- | --- |
| Square | `1024x1024` | Typical fast default |
| Landscape | `1536x1024` | Standard landscape |
| Portrait | `1024x1536` | Standard portrait |
| 2K square | `2048x2048` | Larger square output |
| 2K landscape | `2048x1152` | Widescreen output |
| 4K landscape | `3840x2160` | Widescreen 4K output |
| 4K portrait | `2160x3840` | Vertical 4K output |
| Auto | `auto` | Default size |
Square images are typically fastest to generate. For 4K-style output, use `3840x2160` or `2160x3840`.
## Endpoints
- Generate: `POST /v1/images/generations` (`client.images.generate(...)`)
- Edit: `POST /v1/images/edits` (`client.images.edit(...)`)
## Core parameters for GPT Image models
- `prompt`: text prompt
- `model`: image model
- `n`: number of images (1-10)
- `size`: `auto` by default for `gpt-image-2`; flexible `WIDTHxHEIGHT` sizes are allowed only for `gpt-image-2`; older GPT Image models use `1024x1024`, `1536x1024`, `1024x1536`, or `auto`
- `quality`: `low`, `medium`, `high`, or `auto`
- `background`: output transparency behavior (`transparent`, `opaque`, or `auto`) for generated output; this is not the same thing as the prompt's visual scene/backdrop
- `output_format`: `png` (default), `jpeg`, `webp`
- `output_compression`: 0-100 (jpeg/webp only)
- `moderation`: `auto` (default) or `low`
## Edit-specific parameters
- `image`: one or more input images. For GPT Image models, you can provide up to 16 images.
- `mask`: optional mask image
- `input_fidelity`: `low` or `high` only for models that support it; do not set this for `gpt-image-2`
Model-specific note for `input_fidelity`:
- `gpt-image-2` always uses high fidelity for image inputs and does not support setting `input_fidelity`.
- `gpt-image-1` and `gpt-image-1-mini` preserve all input images, but the first image gets richer textures and finer details.
- `gpt-image-1.5` preserves the first 5 input images with higher fidelity.
## Transparent backgrounds
`gpt-image-2` does not currently support the Image API `background=transparent` parameter. In explicit CLI/API fallback mode, keep `gpt-image-2` when a flat chroma-key background plus local alpha extraction with `python "${CODEX_HOME:-$HOME/.codex}/skills/.system/imagegen/scripts/remove_chroma_key.py"` is acceptable.
Use CLI `gpt-image-1.5` with `background=transparent` and a transparent-capable output format such as `png` or `webp` only after the user explicitly confirms that fallback, unless they already requested `gpt-image-1.5`, `scripts/image_gen.py`, or CLI fallback. If the user asks for true/native transparency, the subject is too complex for clean chroma-key removal, or local background removal fails validation, explain the tradeoff and ask before switching.
## Output
- `data[]` list with `b64_json` per image
- The bundled `scripts/image_gen.py` CLI decodes `b64_json` and writes output files for you.
## Limits and notes
- Input images and masks must be under 50MB.
- Use the edits endpoint when the user requests changes to an existing image.
- Masking is prompt-guided; exact shapes are not guaranteed.
- Large sizes and high quality increase latency and cost.
- Use `quality=low` for fast drafts, thumbnails, and quick iterations. Use `medium` or `high` for final assets, dense text, diagrams, identity-sensitive edits, or high-resolution outputs.
- High `input_fidelity` can materially increase input token usage on models that support it.
- If a request fails because a specific option is unsupported by the selected GPT Image model, retry manually without that option only when the option is not required by the user. If true transparent CLI output is required, ask before switching to `gpt-image-1.5` instead of dropping `background=transparent`, unless the user already explicitly chose that fallback.
## Important boundary
- `quality`, `input_fidelity`, explicit masks, `background`, `output_format`, and related parameters are fallback-only execution controls.
- Do not assume they are built-in `image_gen` tool arguments.

112
.codex-home/skills/.system/imagegen/references/prompting.md Normal file
View file

@ -0,0 +1,112 @@
# Prompting best practices
These prompting principles are shared by both top-level modes of the skill:
- built-in `image_gen` tool (default)
- explicit `scripts/image_gen.py` CLI fallback
This file is about prompt structure, specificity, and iteration. Fallback-only execution controls such as `quality`, `input_fidelity`, masks, output format, and output paths live in the fallback docs.
## Contents
- [Structure](#structure)
- [Specificity policy](#specificity-policy)
- [Allowed and disallowed augmentation](#allowed-and-disallowed-augmentation)
- [Composition and layout](#composition-and-layout)
- [Constraints and invariants](#constraints-and-invariants)
- [Text in images](#text-in-images)
- [Input images and references](#input-images-and-references)
- [Iterate deliberately](#iterate-deliberately)
- [Transparent images](#transparent-images)
- [Fallback-only execution controls](#fallback-only-execution-controls)
- [Use-case tips](#use-case-tips)
- [Where to find copy/paste recipes](#where-to-find-copypaste-recipes)
## Structure
- Use a consistent order: scene/backdrop -> subject -> key details -> constraints -> output intent.
- Include intended use (ad, UI mock, infographic) to set the level of polish.
- For complex requests, use short labeled lines instead of one long paragraph.
## Specificity policy
- If the user prompt is already specific and detailed, normalize it into a clean spec without adding creative requirements.
- If the prompt is generic, you may add tasteful detail when it materially improves the output.
- Treat examples in `sample-prompts.md` as fully-authored recipes, not as the default amount of augmentation to add to every request.
- For photorealism, include `photorealistic` directly when that is the goal, plus concrete real-world texture such as pores, wrinkles, fabric wear, material grain, or imperfect everyday detail.
## Allowed and disallowed augmentation
Allowed augmentation for generic prompts:
- composition and framing cues
- intended-use or polish-level hints
- practical layout guidance
- reasonable scene concreteness that supports the request
Do not add:
- extra characters, props, or objects that are not implied
- brand palettes, slogans, or story beats that are not implied
- arbitrary side-specific placement unless the surrounding layout supports it
## Composition and layout
- Specify framing and viewpoint (close-up, wide, top-down) and placement only when it materially helps.
- Call out negative space if the asset clearly needs room for UI or copy.
- Avoid making left/right layout decisions unless the user or surrounding layout supports them.
- For people, describe body framing, scale, gaze, and object interactions when they matter (`full body visible`, `looking down at the book`, `hands naturally gripping the handlebars`).
## Constraints and invariants
- State what must not change (`keep background unchanged`).
- For edits, say `change only X; keep Y unchanged` and repeat invariants on every iteration to reduce drift.
## Text in images
- Put literal text in quotes or ALL CAPS and specify typography (font style, size, color, placement).
- Spell uncommon words letter-by-letter if accuracy matters.
- For in-image copy, require verbatim rendering and no extra characters.
- In CLI fallback mode, use `medium` or `high` quality for small text, dense infographics, data-heavy slides, multi-font layouts, legends, axes, and footnotes.
## Input images and references
- Do not assume that every provided image is an edit target.
- Label each image by index and role (`Image 1: edit target`, `Image 2: style reference`).
- If the user provides images for style, composition, or mood guidance and does not ask to modify them, treat the request as generation with references.
- If the user asks to preserve an existing image while changing specific parts, treat the request as an edit.
- For compositing, describe how the images interact (`place the subject from Image 2 into Image 1`).
## Iterate deliberately
- Start with a clean base prompt, then make small single-change edits.
- Re-specify critical constraints when you iterate.
- Prefer one targeted follow-up at a time over rewriting the whole prompt.
## Transparent images
- Ask built-in `image_gen` for a genuinely transparent background and preserve its alpha.
## Fallback-only execution controls
- `quality`, `input_fidelity`, explicit masks, output format, and output paths are fallback-only execution controls.
- Do not assume they are built-in `image_gen` tool arguments.
- If the user explicitly chooses CLI fallback, see `references/cli.md` and `references/image-api.md` for those controls.
- In CLI fallback mode, `gpt-image-2` is the default. It supports `quality=low|medium|high|auto`; use `low` for fast drafts and thumbnails, and move to `medium`, `high`, or `auto` for final assets.
- `gpt-image-2` always uses high fidelity for image inputs, so do not set `input_fidelity` with that model.
- CLI `gpt-image-2` does not support `background=transparent`; ask before using `gpt-image-1.5` unless the user explicitly requested that model.
- If the user asks for 4K-style output with `gpt-image-2`, use `3840x2160` for landscape or `2160x3840` for portrait.
## Use-case tips
Generate:
- photorealistic-natural: Prompt as if a real photo is captured in the moment; use photography language (lens, lighting, framing); call for real texture; avoid over-stylized polish unless requested.
- product-mockup: Describe the product/packaging and materials; ensure clean silhouette and label clarity; if in-image text is needed, require verbatim rendering and specify typography.
- ui-mockup: Describe the target fidelity first (shippable mockup or low-fi wireframe), then focus on layout, hierarchy, and practical UI elements; avoid concept-art language.
- infographic-diagram: Define the audience and layout flow; label parts explicitly; require verbatim text; prefer higher quality in CLI mode for dense labels.
- logo-brand: Keep it simple and scalable; ask for a strong silhouette and balanced negative space; avoid decorative flourishes unless requested.
- ads-marketing: Write like a creative brief; include brand positioning, audience, desired vibe, scene, and exact tagline if text must appear.
- productivity-visual: Name the exact artifact (slide, chart, workflow diagram), define the canvas and hierarchy, provide real labels/data, and ask for readable typography and polished spacing.
- scientific-educational: Define audience, lesson objective, required labels, scientific constraints, arrows, and scan-friendly whitespace.
- illustration-story: Define panels or scene beats; keep each action concrete.
- stylized-concept: Specify style cues, material finish, and rendering approach (3D, painterly, clay) without inventing new story elements.
- historical-scene: State the location/date and required period accuracy; constrain clothing, props, and environment to match the era.
Edit:
- text-localization: Change only the text; preserve layout, typography, spacing, and hierarchy; no extra words or reflow unless needed.
- identity-preserve: Lock identity (face, body, pose, hair, expression); change only the specified elements; match lighting and shadows.
- precise-object-edit: Specify exactly what to remove/replace; preserve surrounding texture and lighting; keep everything else unchanged.
- lighting-weather: Change only environmental conditions (light, shadows, atmosphere, precipitation); keep geometry, framing, and subject identity.
- background-extraction: Request a clean cutout on a genuinely transparent background; preserve fine edges and label text; no halos or restyling.
- style-transfer: Specify style cues to preserve (palette, texture, brushwork) and what must change; add `no extra elements` to prevent drift.
- compositing: Reference inputs by index; specify what moves where; match lighting, perspective, and scale; keep the base framing unchanged.
- sketch-to-render: Preserve layout, proportions, and perspective; choose materials and lighting that support the supplied sketch without adding new elements.
## Where to find copy/paste recipes
For copy/paste prompt specs (examples only), see `references/sample-prompts.md`. This file focuses on principles, specificity, and iteration patterns.

422
.codex-home/skills/.system/imagegen/references/sample-prompts.md Normal file
View file

@ -0,0 +1,422 @@
# Sample prompts (copy/paste)
These prompt recipes are shared across both top-level modes of the skill:
- built-in `image_gen` tool (default)
- `scripts/image_gen.py` CLI fallback for explicit or user-confirmed CLI/API/model requests
Use these as starting points. They are intentionally complete prompt recipes, not the default amount of augmentation to add to every user request.
When adapting a user's prompt:
- keep user-provided requirements
- only add detail according to the specificity policy in `SKILL.md`
- do not treat every example below as permission to invent extra story elements
The labeled lines are prompt scaffolding, not a closed schema. `Asset type` and `Input images` are prompt-only scaffolding; the CLI does not expose them as dedicated flags.
Execution details such as explicit CLI flags, `quality`, `input_fidelity`, masks, output formats, and local output paths depend on mode. Use built-in `image_gen` by default, request transparent backgrounds directly, and preserve the generated alpha; apply CLI-specific controls only when the user chooses or confirms that fallback.
CLI model notes:
- `gpt-image-2` is the fallback CLI default for new workflows.
- `gpt-image-2` supports `quality` values `low`, `medium`, `high`, and `auto`.
- For 4K-style `gpt-image-2` output, use `3840x2160` or `2160x3840`.
- CLI `gpt-image-2` does not support `background=transparent`; ask before using `gpt-image-1.5` unless the user explicitly requested that model.
- Do not set `input_fidelity` with `gpt-image-2`; image inputs already use high fidelity.
For prompting principles (structure, specificity, invariants, iteration), see `references/prompting.md`.
## Generate
### photorealistic-natural
```
Use case: photorealistic-natural
Primary request: candid photo of an elderly sailor on a small fishing boat adjusting a net
Scene/backdrop: coastal water with soft haze
Subject: weathered skin with wrinkles and sun texture
Style/medium: photorealistic candid photo
Composition/framing: medium close-up, eye-level
Lighting/mood: soft coastal daylight, shallow depth of field, subtle film grain
Materials/textures: real skin texture, worn fabric, salt-worn wood
Constraints: natural color balance; no heavy retouching; no glamorization; no watermark
Avoid: studio polish; staged look
```
### product-mockup
```
Use case: product-mockup
Primary request: premium product photo of a matte black shampoo bottle with a minimal label
Scene/backdrop: clean studio gradient from light gray to white
Subject: single bottle centered with subtle reflection
Style/medium: premium product photography
Composition/framing: centered, slight three-quarter angle, generous padding
Lighting/mood: softbox lighting, clean highlights, controlled shadows
Materials/textures: matte plastic, crisp label printing
Constraints: no logos or trademarks; no watermark
```
### ui-mockup
```
Use case: ui-mockup
Primary request: mobile app home screen for a local farmers market with vendors and daily specials
Asset type: mobile app screen
Style/medium: realistic product UI, not concept art
Composition/framing: clean vertical mobile layout with clear hierarchy
Constraints: practical layout, clear typography, no logos or trademarks, no watermark
```
### infographic-diagram
```
Use case: infographic-diagram
Primary request: detailed infographic of an automatic coffee machine flow
Scene/backdrop: clean, light neutral background
Subject: bean hopper -> grinder -> brew group -> boiler -> water tank -> drip tray
Style/medium: clean vector-like infographic with clear callouts and arrows
Composition/framing: vertical poster layout, top-to-bottom flow
Text (verbatim): "Bean Hopper", "Grinder", "Brew Group", "Boiler", "Water Tank", "Drip Tray"
Constraints: clear labels, strong contrast, no logos or trademarks, no watermark
```
### scientific-educational
```
Use case: scientific-educational
Primary request: biology diagram titled "Cellular Respiration at a Glance" for high school students
Scene/backdrop: clean white classroom handout background
Subject: glucose turns into energy inside a cell; include glycolysis, Krebs cycle, and electron transport chain
Style/medium: flat scientific diagram with consistent icons, arrows, and readable labels
Composition/framing: landscape slide-style layout with clear hierarchy and generous whitespace
Text (verbatim): "Cellular Respiration at a Glance", "Glucose", "Pyruvate", "ATP", "NADH", "FADH2", "CO2", "O2", "H2O"
Constraints: scientifically plausible; avoid tiny text; no extra decoration; no watermark
```
### logo-brand
```
Use case: logo-brand
Primary request: original logo for "Field & Flour", a local bakery
Style/medium: vector logo mark; flat colors; minimal
Composition/framing: single centered logo on a plain background with generous padding
Constraints: strong silhouette, balanced negative space; original design only; no gradients unless essential; no trademarks; no watermark
```
### illustration-story
```
Use case: illustration-story
Primary request: 4-panel comic about a pet left alone at home
Scene/backdrop: cozy living room across panels
Subject: pet reacting to the owner leaving, then relaxing, then returning to a composed pose
Style/medium: comic illustration with clear panels
Composition/framing: 4 equal-sized vertical panels, readable actions per panel
Constraints: no text; no logos or trademarks; no watermark
```
### stylized-concept
```
Use case: stylized-concept
Primary request: cavernous hangar interior with tall support beams and drifting fog
Scene/backdrop: industrial hangar interior, deep scale, light haze
Subject: compact shuttle parked near the center
Style/medium: cinematic concept art, industrial realism
Composition/framing: wide-angle, low-angle
Lighting/mood: volumetric light rays cutting through fog
Constraints: no logos or trademarks; no watermark
```
### ads-marketing
```
Use case: ads-marketing
Primary request: campaign image for a streetwear brand called Thread
Subject: group of friends hanging out together in a stylish urban setting
Style/medium: polished youth streetwear campaign photography
Composition/framing: vertical ad layout with natural poses and integrated headline space
Lighting/mood: contemporary, energetic, tasteful
Text (verbatim): "Yours to Create."
Constraints: render the tagline exactly once; clean legible typography; no extra text; no watermarks; no unrelated logos
```
### productivity-visual
```
Use case: productivity-visual
Primary request: one pitch-deck slide titled "Market Opportunity"
Asset type: fundraising slide image
Style/medium: clean modern deck slide, white background, crisp sans-serif typography
Subject: TAM/SAM/SOM concentric-circle diagram plus a small growth bar chart from 2021 to 2026
Composition/framing: 16:9 landscape slide, clear data hierarchy, polished spacing
Text (verbatim): "Market Opportunity", "TAM: $42B", "SAM: $8.7B", "SOM: $340M", "AGI Research, 2024", "Internal analysis"
Constraints: readable labels, no clip art, no stock photography, no decorative clutter, no watermark
```
### historical-scene
```
Use case: historical-scene
Primary request: outdoor crowd scene in Bethel, New York on August 16, 1969
Scene/backdrop: open field with period-appropriate staging
Subject: crowd in period-accurate clothing, authentic environment
Style/medium: photorealistic photo
Composition/framing: wide shot, eye-level
Constraints: period-accurate details; no modern objects; no logos or trademarks; no watermark
```
## Asset type templates (taxonomy-aligned)
### Website assets template
```
Use case: <photorealistic-natural|stylized-concept|product-mockup|infographic-diagram|ui-mockup>
Asset type: <hero image / section illustration / blog header>
Primary request: <short description>
Scene/backdrop: <environment or abstract backdrop>
Subject: <main subject>
Style/medium: <photo/illustration/3D>
Composition/framing: <wide/centered; note usable negative space only if needed>
Lighting/mood: <soft/bright/neutral>
Color palette: <brand colors or neutral>
Constraints: <no text; no logos; no watermark; leave room for UI if needed>
```
### Website assets example: minimal hero background
```
Use case: stylized-concept
Asset type: landing page hero background
Primary request: minimal abstract background with a soft gradient and subtle texture
Style/medium: matte illustration / soft-rendered abstract background
Composition/framing: wide composition with usable negative space for page copy
Lighting/mood: gentle studio glow
Color palette: restrained neutral palette
Constraints: no text; no logos; no watermark
```
### Website assets example: feature section illustration
```
Use case: stylized-concept
Asset type: feature section illustration
Primary request: simple abstract shapes suggesting connection and flow
Scene/backdrop: subtle light-gray backdrop with faint texture
Style/medium: flat illustration; soft shadows; restrained contrast
Composition/framing: centered cluster; open margins for UI
Color palette: muted neutral palette
Constraints: no text; no logos; no watermark
```
### Website assets example: blog header image
```
Use case: photorealistic-natural
Asset type: blog header image
Primary request: overhead desk scene with notebook, pen, and coffee cup
Scene/backdrop: warm wooden tabletop
Style/medium: photorealistic photo
Composition/framing: wide crop with clean room for page copy
Lighting/mood: soft morning light
Constraints: no text; no logos; no watermark
```
### Game assets template
```
Use case: stylized-concept
Asset type: <game environment concept art / game character concept / game UI icon / tileable game texture>
Primary request: <biome/scene/character/icon/material>
Scene/backdrop: <location + set dressing> (if applicable)
Subject: <main focal element(s)>
Style/medium: <realistic/stylized>; <concept art / character render / UI icon / texture>
Composition/framing: <wide/establishing/top-down>; <camera angle>; <focal point placement>
Lighting/mood: <time of day>; <mood>; <volumetric/fog/etc>
Constraints: no logos or trademarks; no watermark
```
### Game assets example: environment concept art
```
Use case: stylized-concept
Asset type: game environment concept art
Primary request: cavernous hangar interior with tall support beams and drifting fog
Scene/backdrop: industrial hangar interior, deep scale, light haze
Subject: compact shuttle parked near the center
Style/medium: cinematic concept art, industrial realism
Composition/framing: wide-angle, low-angle
Lighting/mood: volumetric light rays cutting through fog
Constraints: no logos or trademarks; no watermark
```
### Game assets example: character concept
```
Use case: stylized-concept
Asset type: game character concept
Primary request: desert scout character with layered travel gear
Subject: long coat, satchel, practical travel clothing
Style/medium: character render; stylized realism
Composition/framing: neutral hero pose on a simple backdrop
Constraints: no logos or trademarks; no watermark
```
### Game assets example: UI icon
```
Use case: stylized-concept
Asset type: game UI icon
Primary request: round shield icon with a subtle rune pattern
Style/medium: painted game UI icon
Composition/framing: centered icon; generous padding; clear silhouette
Constraints: no text; no background scene elements; no logos or trademarks; no watermark
```
### Game assets example: tileable texture
```
Use case: stylized-concept
Asset type: tileable game texture
Primary request: worn sandstone blocks
Style/medium: seamless tileable texture; PBR-ish look
Scene/backdrop: neutral lighting reference only
Constraints: seamless edges; no obvious focal elements; no text; no logos or trademarks; no watermark
```
### Wireframe template
```
Use case: ui-mockup
Asset type: website wireframe
Primary request: <page or flow to sketch>
Style/medium: low-fi grayscale wireframe
Composition/framing: <landscape or portrait to match expected device>
Subject: <sections in order; grid/columns; key labels>
Constraints: no color; no logos; no real photos; no watermark
```
### Wireframe example: homepage (desktop)
```
Use case: ui-mockup
Asset type: website wireframe
Primary request: SaaS homepage layout with clear hierarchy
Style/medium: low-fi grayscale wireframe
Subject: top nav; hero with headline and CTA; three feature cards; testimonial strip; pricing preview; footer
Composition/framing: landscape desktop layout
Constraints: label major blocks; no color; no logos; no real photos; no watermark
```
### Wireframe example: pricing page
```
Use case: ui-mockup
Asset type: website wireframe
Primary request: pricing page layout with comparison table
Style/medium: low-fi grayscale wireframe
Subject: header; plan toggle; 3 pricing cards; comparison table; FAQ accordion; footer
Composition/framing: desktop or tablet layout
Constraints: label key areas; no color; no logos; no real photos; no watermark
```
### Wireframe example: mobile onboarding flow
```
Use case: ui-mockup
Asset type: mobile onboarding wireframe
Primary request: three-screen mobile onboarding flow
Style/medium: low-fi grayscale wireframe
Subject: screen 1 headline and CTA; screen 2 feature bullets; screen 3 form fields and CTA
Composition/framing: portrait mobile layout
Constraints: label screens and blocks; no color; no logos; no real photos; no watermark
```
### Logo template
```
Use case: logo-brand
Asset type: logo concept
Primary request: <brand idea or symbol concept>
Style/medium: vector logo mark; flat colors; minimal
Composition/framing: centered mark; clear silhouette; generous margin
Color palette: <1-2 colors; high contrast>
Text (verbatim): "<exact name>" (only if needed)
Constraints: no gradients; no mockups; no 3D; no watermark
```
### Logo example: abstract symbol mark
```
Use case: logo-brand
Asset type: logo concept
Primary request: geometric leaf symbol suggesting sustainability and growth
Style/medium: vector logo mark; flat colors; minimal
Composition/framing: centered mark; clear silhouette
Color palette: deep green and off-white
Constraints: no text unless requested; no gradients; no mockups; no 3D; no watermark
```
### Logo example: monogram mark
```
Use case: logo-brand
Asset type: logo concept
Primary request: interlocking monogram of the letters "AV"
Style/medium: vector logo mark; flat colors; minimal
Composition/framing: centered mark; balanced spacing
Color palette: black on white
Constraints: no gradients; no mockups; no 3D; no watermark
```
### Logo example: wordmark
```
Use case: logo-brand
Asset type: logo concept
Primary request: clean wordmark for a modern studio
Style/medium: vector wordmark; flat colors; minimal
Text (verbatim): "Studio North"
Composition/framing: centered text; even letter spacing
Constraints: no gradients; no mockups; no 3D; no watermark
```
## Edit
### text-localization
```
Use case: text-localization
Input images: Image 1: original infographic
Primary request: replace "Bean Hopper", "Grinder", "Brew Group", "Boiler", "Water Tank", and "Drip Tray" with "Tolva", "Molino", "Grupo de infusión", "Caldera", "Depósito de agua", and "Bandeja de goteo"
Constraints: change only the text; preserve layout, typography, spacing, and hierarchy; no extra words; do not alter logos or imagery
```
### identity-preserve
```
Use case: identity-preserve
Input images: Image 1: person photo; Image 2..N: clothing references
Primary request: replace only the clothing with the provided garments
Constraints: preserve face, body shape, pose, hair, expression, and identity; match lighting and shadows; keep the background unchanged; no accessories or text
```
### precise-object-edit
```
Use case: precise-object-edit
Input images: Image 1: room photo
Primary request: replace only the white chairs with wooden chairs
Constraints: preserve camera angle, room lighting, floor shadows, and surrounding objects; keep all other aspects unchanged
```
### lighting-weather
```
Use case: lighting-weather
Input images: Image 1: original photo
Primary request: make it look like a winter evening with gentle snowfall
Constraints: preserve subject identity, geometry, camera angle, and composition; change only lighting, atmosphere, and weather
```
### style-transfer
```
Use case: style-transfer
Input images: Image 1: style reference
Primary request: apply Image 1's visual style to a man riding a motorcycle on a plain white backdrop
Constraints: preserve palette, texture, and brushwork; no extra elements
```
### compositing
```
Use case: compositing
Input images: Image 1: base scene; Image 2: subject to insert
Primary request: place the subject from Image 2 next to the person in Image 1
Constraints: match lighting, perspective, and scale; keep the base framing unchanged; no extra elements
```
### character consistency workflow
```
Use case: identity-preserve
Input images: Image 1: previous character anchor illustration
Primary request: continue the story with the same character in a new scene and action
Scene/backdrop: snowy forest after a winter storm
Subject: same young forest hero gently helping a frightened squirrel out of a fallen tree
Style/medium: same children's book watercolor illustration style as Image 1
Constraints: do not redesign the character; preserve facial features, proportions, outfit, color palette, and personality; no text; no watermark
```
### sketch-to-render
```
Use case: sketch-to-render
Input images: Image 1: drawing
Primary request: turn the drawing into a photorealistic image
Constraints: preserve layout, proportions, and perspective; choose realistic materials and lighting; do not add new elements or text
```

995
.codex-home/skills/.system/imagegen/scripts/image_gen.py Normal file
View file

@ -0,0 +1,995 @@
#!/usr/bin/env python3
"""Fallback CLI for explicit image generation or editing with GPT Image models.
Used only when the user explicitly opts into CLI fallback mode, or when explicit
transparent output requires the `gpt-image-1.5` fallback path.
Defaults to gpt-image-2 and a structured prompt augmentation workflow.
"""
from __future__ import annotations
import argparse
import asyncio
import base64
import json
import os
from pathlib import Path
import re
import sys
import time
from typing import Any, Dict, Iterable, List, Optional, Tuple
from io import BytesIO
DEFAULT_MODEL = "gpt-image-2"
DEFAULT_SIZE = "auto"
DEFAULT_QUALITY = "medium"
DEFAULT_OUTPUT_FORMAT = "png"
DEFAULT_CONCURRENCY = 5
DEFAULT_DOWNSCALE_SUFFIX = "-web"
DEFAULT_OUTPUT_PATH = "output/imagegen/output.png"
GPT_IMAGE_MODEL_PREFIX = "gpt-image-"
ALLOWED_LEGACY_SIZES = {"1024x1024", "1536x1024", "1024x1536", "auto"}
ALLOWED_QUALITIES = {"low", "medium", "high", "auto"}
ALLOWED_BACKGROUNDS = {"transparent", "opaque", "auto", None}
ALLOWED_INPUT_FIDELITIES = {"low", "high", None}
GPT_IMAGE_2_MODEL = "gpt-image-2"
GPT_IMAGE_2_MIN_PIXELS = 655_360
GPT_IMAGE_2_MAX_PIXELS = 8_294_400
GPT_IMAGE_2_MAX_EDGE = 3840
GPT_IMAGE_2_MAX_RATIO = 3.0
MAX_IMAGE_BYTES = 50 * 1024 * 1024
MAX_BATCH_JOBS = 500
def _die(message: str, code: int = 1) -> None:
print(f"Error: {message}", file=sys.stderr)
raise SystemExit(code)
def _warn(message: str) -> None:
print(f"Warning: {message}", file=sys.stderr)
def _dependency_hint(package: str, *, upgrade: bool = False) -> str:
command = f"uv pip install {'-U ' if upgrade else ''}{package}"
return (
"Activate the repo-selected environment first, then install it with "
f"`{command}`. If this repo uses a local virtualenv, start with "
"`source .venv/bin/activate`; otherwise use this repo's configured shared fallback "
"environment. If your project declares dependencies, prefer that project's normal "
"`uv sync` flow."
)
def _ensure_api_key(dry_run: bool) -> None:
if os.getenv("OPENAI_API_KEY"):
print("OPENAI_API_KEY is set.", file=sys.stderr)
return
if dry_run:
_warn("OPENAI_API_KEY is not set; dry-run only.")
return
_die("OPENAI_API_KEY is not set. Export it before running.")
def _read_prompt(prompt: Optional[str], prompt_file: Optional[str]) -> str:
if prompt and prompt_file:
_die("Use --prompt or --prompt-file, not both.")
if prompt_file:
path = Path(prompt_file)
if not path.exists():
_die(f"Prompt file not found: {path}")
return path.read_text(encoding="utf-8").strip()
if prompt:
return prompt.strip()
_die("Missing prompt. Use --prompt or --prompt-file.")
return "" # unreachable
def _check_image_paths(paths: Iterable[str]) -> List[Path]:
resolved: List[Path] = []
for raw in paths:
path = Path(raw)
if not path.exists():
_die(f"Image file not found: {path}")
if path.stat().st_size > MAX_IMAGE_BYTES:
_warn(f"Image exceeds 50MB limit: {path}")
resolved.append(path)
return resolved
def _normalize_output_format(fmt: Optional[str]) -> str:
if not fmt:
return DEFAULT_OUTPUT_FORMAT
fmt = fmt.lower()
if fmt not in {"png", "jpeg", "jpg", "webp"}:
_die("output-format must be png, jpeg, jpg, or webp.")
return "jpeg" if fmt == "jpg" else fmt
def _parse_size(size: str) -> Optional[Tuple[int, int]]:
match = re.fullmatch(r"([1-9][0-9]*)x([1-9][0-9]*)", size)
if not match:
return None
return int(match.group(1)), int(match.group(2))
def _validate_gpt_image_2_size(size: str) -> None:
if size == "auto":
return
parsed = _parse_size(size)
if parsed is None:
_die("size must be auto or WIDTHxHEIGHT, for example 1024x1024.")
width, height = parsed
max_edge = max(width, height)
min_edge = min(width, height)
total_pixels = width * height
if max_edge > GPT_IMAGE_2_MAX_EDGE:
_die("gpt-image-2 size maximum edge length must be less than or equal to 3840px.")
if width % 16 != 0 or height % 16 != 0:
_die("gpt-image-2 size width and height must be multiples of 16px.")
if max_edge / min_edge > GPT_IMAGE_2_MAX_RATIO:
_die("gpt-image-2 size long edge to short edge ratio must not exceed 3:1.")
if total_pixels < GPT_IMAGE_2_MIN_PIXELS or total_pixels > GPT_IMAGE_2_MAX_PIXELS:
_die(
"gpt-image-2 size total pixels must be at least 655,360 and no more than 8,294,400."
)
def _validate_size(size: str, model: str) -> None:
if model == GPT_IMAGE_2_MODEL:
_validate_gpt_image_2_size(size)
return
if size not in ALLOWED_LEGACY_SIZES:
_die(
"size must be one of 1024x1024, 1536x1024, 1024x1536, or auto for this GPT Image model."
)
def _validate_quality(quality: str) -> None:
if quality not in ALLOWED_QUALITIES:
_die("quality must be one of low, medium, high, or auto.")
def _validate_background(background: Optional[str]) -> None:
if background not in ALLOWED_BACKGROUNDS:
_die("background must be one of transparent, opaque, or auto.")
def _validate_input_fidelity(input_fidelity: Optional[str]) -> None:
if input_fidelity not in ALLOWED_INPUT_FIDELITIES:
_die("input-fidelity must be one of low or high.")
def _validate_model(model: str) -> None:
if not model.startswith(GPT_IMAGE_MODEL_PREFIX):
_die(
"model must be a GPT Image model (for example gpt-image-1.5, gpt-image-1, or gpt-image-1-mini)."
)
def _validate_transparency(background: Optional[str], output_format: str) -> None:
if background == "transparent" and output_format not in {"png", "webp"}:
_die("transparent background requires output-format png or webp.")
def _validate_model_specific_options(
*,
model: str,
background: Optional[str],
input_fidelity: Optional[str] = None,
) -> None:
if model != GPT_IMAGE_2_MODEL:
return
if background == "transparent":
_die(
"transparent backgrounds are not supported in gpt-image-2, the latest model. "
"Use --model gpt-image-1.5 --background transparent --output-format png instead."
)
if input_fidelity is not None:
_die(
"input_fidelity is not supported in gpt-image-2 because image inputs always use high fidelity for this model."
)
def _validate_generate_payload(payload: Dict[str, Any]) -> None:
model = str(payload.get("model", DEFAULT_MODEL))
_validate_model(model)
n = int(payload.get("n", 1))
if n < 1 or n > 10:
_die("n must be between 1 and 10")
size = str(payload.get("size", DEFAULT_SIZE))
quality = str(payload.get("quality", DEFAULT_QUALITY))
background = payload.get("background")
_validate_size(size, model)
_validate_quality(quality)
_validate_background(background)
_validate_model_specific_options(model=model, background=background)
oc = payload.get("output_compression")
if oc is not None and not (0 <= int(oc) <= 100):
_die("output_compression must be between 0 and 100")
def _build_output_paths(
out: str,
output_format: str,
count: int,
out_dir: Optional[str],
) -> List[Path]:
ext = "." + output_format
if out_dir:
out_base = Path(out_dir)
out_base.mkdir(parents=True, exist_ok=True)
return [out_base / f"image_{i}{ext}" for i in range(1, count + 1)]
out_path = Path(out)
if out_path.exists() and out_path.is_dir():
out_path.mkdir(parents=True, exist_ok=True)
return [out_path / f"image_{i}{ext}" for i in range(1, count + 1)]
if out_path.suffix == "":
out_path = out_path.with_suffix(ext)
elif output_format and out_path.suffix.lstrip(".").lower() != output_format:
_warn(
f"Output extension {out_path.suffix} does not match output-format {output_format}."
)
if count == 1:
return [out_path]
return [
out_path.with_name(f"{out_path.stem}-{i}{out_path.suffix}")
for i in range(1, count + 1)
]
def _augment_prompt(args: argparse.Namespace, prompt: str) -> str:
fields = _fields_from_args(args)
return _augment_prompt_fields(args.augment, prompt, fields)
def _augment_prompt_fields(augment: bool, prompt: str, fields: Dict[str, Optional[str]]) -> str:
if not augment:
return prompt
sections: List[str] = []
if fields.get("use_case"):
sections.append(f"Use case: {fields['use_case']}")
sections.append(f"Primary request: {prompt}")
if fields.get("scene"):
sections.append(f"Scene/background: {fields['scene']}")
if fields.get("subject"):
sections.append(f"Subject: {fields['subject']}")
if fields.get("style"):
sections.append(f"Style/medium: {fields['style']}")
if fields.get("composition"):
sections.append(f"Composition/framing: {fields['composition']}")
if fields.get("lighting"):
sections.append(f"Lighting/mood: {fields['lighting']}")
if fields.get("palette"):
sections.append(f"Color palette: {fields['palette']}")
if fields.get("materials"):
sections.append(f"Materials/textures: {fields['materials']}")
if fields.get("text"):
sections.append(f"Text (verbatim): \"{fields['text']}\"")
if fields.get("constraints"):
sections.append(f"Constraints: {fields['constraints']}")
if fields.get("negative"):
sections.append(f"Avoid: {fields['negative']}")
return "\n".join(sections)
def _fields_from_args(args: argparse.Namespace) -> Dict[str, Optional[str]]:
return {
"use_case": getattr(args, "use_case", None),
"scene": getattr(args, "scene", None),
"subject": getattr(args, "subject", None),
"style": getattr(args, "style", None),
"composition": getattr(args, "composition", None),
"lighting": getattr(args, "lighting", None),
"palette": getattr(args, "palette", None),
"materials": getattr(args, "materials", None),
"text": getattr(args, "text", None),
"constraints": getattr(args, "constraints", None),
"negative": getattr(args, "negative", None),
}
def _print_request(payload: dict) -> None:
print(json.dumps(payload, indent=2, sort_keys=True))
def _decode_and_write(images: List[str], outputs: List[Path], force: bool) -> None:
for idx, image_b64 in enumerate(images):
if idx >= len(outputs):
break
out_path = outputs[idx]
if out_path.exists() and not force:
_die(f"Output already exists: {out_path} (use --force to overwrite)")
out_path.parent.mkdir(parents=True, exist_ok=True)
out_path.write_bytes(base64.b64decode(image_b64))
print(f"Wrote {out_path}")
def _derive_downscale_path(path: Path, suffix: str) -> Path:
if suffix and not suffix.startswith("-") and not suffix.startswith("_"):
suffix = "-" + suffix
return path.with_name(f"{path.stem}{suffix}{path.suffix}")
def _downscale_image_bytes(image_bytes: bytes, *, max_dim: int, output_format: str) -> bytes:
try:
from PIL import Image
except Exception:
_die(f"Downscaling requires Pillow. {_dependency_hint('pillow')}")
if max_dim < 1:
_die("--downscale-max-dim must be >= 1")
with Image.open(BytesIO(image_bytes)) as img:
img.load()
w, h = img.size
scale = min(1.0, float(max_dim) / float(max(w, h)))
target = (max(1, int(round(w * scale))), max(1, int(round(h * scale))))
resized = img if target == (w, h) else img.resize(target, Image.Resampling.LANCZOS)
fmt = output_format.lower()
if fmt == "jpg":
fmt = "jpeg"
if fmt == "jpeg":
if resized.mode in ("RGBA", "LA") or ("transparency" in getattr(resized, "info", {})):
bg = Image.new("RGB", resized.size, (255, 255, 255))
bg.paste(resized.convert("RGBA"), mask=resized.convert("RGBA").split()[-1])
resized = bg
else:
resized = resized.convert("RGB")
out = BytesIO()
resized.save(out, format=fmt.upper())
return out.getvalue()
def _decode_write_and_downscale(
images: List[str],
outputs: List[Path],
*,
force: bool,
downscale_max_dim: Optional[int],
downscale_suffix: str,
output_format: str,
) -> None:
for idx, image_b64 in enumerate(images):
if idx >= len(outputs):
break
out_path = outputs[idx]
if out_path.exists() and not force:
_die(f"Output already exists: {out_path} (use --force to overwrite)")
out_path.parent.mkdir(parents=True, exist_ok=True)
raw = base64.b64decode(image_b64)
out_path.write_bytes(raw)
print(f"Wrote {out_path}")
if downscale_max_dim is None:
continue
derived = _derive_downscale_path(out_path, downscale_suffix)
if derived.exists() and not force:
_die(f"Output already exists: {derived} (use --force to overwrite)")
derived.parent.mkdir(parents=True, exist_ok=True)
resized = _downscale_image_bytes(raw, max_dim=downscale_max_dim, output_format=output_format)
derived.write_bytes(resized)
print(f"Wrote {derived}")
def _create_client():
try:
from openai import OpenAI
except ImportError:
_die(f"openai SDK not installed in the active environment. {_dependency_hint('openai')}")
return OpenAI()
def _create_async_client():
try:
from openai import AsyncOpenAI
except ImportError:
try:
import openai as _openai # noqa: F401
except ImportError:
_die(
f"openai SDK not installed in the active environment. {_dependency_hint('openai')}"
)
_die(
"AsyncOpenAI not available in this openai SDK version. "
f"{_dependency_hint('openai', upgrade=True)}"
)
return AsyncOpenAI()
def _slugify(value: str) -> str:
value = value.strip().lower()
value = re.sub(r"[^a-z0-9]+", "-", value)
value = re.sub(r"-{2,}", "-", value).strip("-")
return value[:60] if value else "job"
def _normalize_job(job: Any, idx: int) -> Dict[str, Any]:
if isinstance(job, str):
prompt = job.strip()
if not prompt:
_die(f"Empty prompt at job {idx}")
return {"prompt": prompt}
if isinstance(job, dict):
if "prompt" not in job or not str(job["prompt"]).strip():
_die(f"Missing prompt for job {idx}")
return job
_die(f"Invalid job at index {idx}: expected string or object.")
return {} # unreachable
def _read_jobs_jsonl(path: str) -> List[Dict[str, Any]]:
p = Path(path)
if not p.exists():
_die(f"Input file not found: {p}")
jobs: List[Dict[str, Any]] = []
for line_no, raw in enumerate(p.read_text(encoding="utf-8").splitlines(), start=1):
line = raw.strip()
if not line or line.startswith("#"):
continue
try:
item: Any
if line.startswith("{"):
item = json.loads(line)
else:
item = line
jobs.append(_normalize_job(item, idx=line_no))
except json.JSONDecodeError as exc:
_die(f"Invalid JSON on line {line_no}: {exc}")
if not jobs:
_die("No jobs found in input file.")
if len(jobs) > MAX_BATCH_JOBS:
_die(f"Too many jobs ({len(jobs)}). Max is {MAX_BATCH_JOBS}.")
return jobs
def _merge_non_null(dst: Dict[str, Any], src: Dict[str, Any]) -> Dict[str, Any]:
merged = dict(dst)
for k, v in src.items():
if v is not None:
merged[k] = v
return merged
def _job_output_paths(
*,
out_dir: Path,
output_format: str,
idx: int,
prompt: str,
n: int,
explicit_out: Optional[str],
) -> List[Path]:
out_dir.mkdir(parents=True, exist_ok=True)
ext = "." + output_format
if explicit_out:
base = Path(explicit_out)
if base.suffix == "":
base = base.with_suffix(ext)
elif base.suffix.lstrip(".").lower() != output_format:
_warn(
f"Job {idx}: output extension {base.suffix} does not match output-format {output_format}."
)
base = out_dir / base.name
else:
slug = _slugify(prompt[:80])
base = out_dir / f"{idx:03d}-{slug}{ext}"
if n == 1:
return [base]
return [
base.with_name(f"{base.stem}-{i}{base.suffix}")
for i in range(1, n + 1)
]
def _extract_retry_after_seconds(exc: Exception) -> Optional[float]:
# Best-effort: openai SDK errors vary by version. Prefer a conservative fallback.
for attr in ("retry_after", "retry_after_seconds"):
val = getattr(exc, attr, None)
if isinstance(val, (int, float)) and val >= 0:
return float(val)
msg = str(exc)
m = re.search(r"retry[- ]after[:= ]+([0-9]+(?:\\.[0-9]+)?)", msg, re.IGNORECASE)
if m:
try:
return float(m.group(1))
except Exception:
return None
return None
def _is_rate_limit_error(exc: Exception) -> bool:
name = exc.__class__.__name__.lower()
if "ratelimit" in name or "rate_limit" in name:
return True
msg = str(exc).lower()
return "429" in msg or "rate limit" in msg or "too many requests" in msg
def _is_transient_error(exc: Exception) -> bool:
if _is_rate_limit_error(exc):
return True
name = exc.__class__.__name__.lower()
if "timeout" in name or "timedout" in name or "tempor" in name:
return True
msg = str(exc).lower()
return "timeout" in msg or "timed out" in msg or "connection reset" in msg
async def _generate_one_with_retries(
client: Any,
payload: Dict[str, Any],
*,
attempts: int,
job_label: str,
) -> Any:
last_exc: Optional[Exception] = None
for attempt in range(1, attempts + 1):
try:
return await client.images.generate(**payload)
except Exception as exc:
last_exc = exc
if not _is_transient_error(exc):
raise
if attempt == attempts:
raise
sleep_s = _extract_retry_after_seconds(exc)
if sleep_s is None:
sleep_s = min(60.0, 2.0**attempt)
print(
f"{job_label} attempt {attempt}/{attempts} failed ({exc.__class__.__name__}); retrying in {sleep_s:.1f}s",
file=sys.stderr,
)
await asyncio.sleep(sleep_s)
raise last_exc or RuntimeError("unknown error")
async def _run_generate_batch(args: argparse.Namespace) -> int:
jobs = _read_jobs_jsonl(args.input)
out_dir = Path(args.out_dir)
base_fields = _fields_from_args(args)
base_payload = {
"model": args.model,
"n": args.n,
"size": args.size,
"quality": args.quality,
"background": args.background,
"output_format": args.output_format,
"output_compression": args.output_compression,
"moderation": args.moderation,
}
if args.dry_run:
for i, job in enumerate(jobs, start=1):
prompt = str(job["prompt"]).strip()
fields = _merge_non_null(base_fields, job.get("fields", {}))
# Allow flat job keys as well (use_case, scene, etc.)
fields = _merge_non_null(fields, {k: job.get(k) for k in base_fields.keys()})
augmented = _augment_prompt_fields(args.augment, prompt, fields)
job_payload = dict(base_payload)
job_payload["prompt"] = augmented
job_payload = _merge_non_null(job_payload, {k: job.get(k) for k in base_payload.keys()})
job_payload = {k: v for k, v in job_payload.items() if v is not None}
_validate_generate_payload(job_payload)
effective_output_format = _normalize_output_format(job_payload.get("output_format"))
_validate_transparency(job_payload.get("background"), effective_output_format)
job_payload["output_format"] = effective_output_format
n = int(job_payload.get("n", 1))
outputs = _job_output_paths(
out_dir=out_dir,
output_format=effective_output_format,
idx=i,
prompt=prompt,
n=n,
explicit_out=job.get("out"),
)
downscaled = None
if args.downscale_max_dim is not None:
downscaled = [
str(_derive_downscale_path(p, args.downscale_suffix)) for p in outputs
]
_print_request(
{
"endpoint": "/v1/images/generations",
"job": i,
"outputs": [str(p) for p in outputs],
"outputs_downscaled": downscaled,
**job_payload,
}
)
return 0
client = _create_async_client()
sem = asyncio.Semaphore(args.concurrency)
any_failed = False
async def run_job(i: int, job: Dict[str, Any]) -> Tuple[int, Optional[str]]:
nonlocal any_failed
prompt = str(job["prompt"]).strip()
job_label = f"[job {i}/{len(jobs)}]"
fields = _merge_non_null(base_fields, job.get("fields", {}))
fields = _merge_non_null(fields, {k: job.get(k) for k in base_fields.keys()})
augmented = _augment_prompt_fields(args.augment, prompt, fields)
payload = dict(base_payload)
payload["prompt"] = augmented
payload = _merge_non_null(payload, {k: job.get(k) for k in base_payload.keys()})
payload = {k: v for k, v in payload.items() if v is not None}
n = int(payload.get("n", 1))
_validate_generate_payload(payload)
effective_output_format = _normalize_output_format(payload.get("output_format"))
_validate_transparency(payload.get("background"), effective_output_format)
payload["output_format"] = effective_output_format
outputs = _job_output_paths(
out_dir=out_dir,
output_format=effective_output_format,
idx=i,
prompt=prompt,
n=n,
explicit_out=job.get("out"),
)
try:
async with sem:
print(f"{job_label} starting", file=sys.stderr)
started = time.time()
result = await _generate_one_with_retries(
client,
payload,
attempts=args.max_attempts,
job_label=job_label,
)
elapsed = time.time() - started
print(f"{job_label} completed in {elapsed:.1f}s", file=sys.stderr)
images = [item.b64_json for item in result.data]
_decode_write_and_downscale(
images,
outputs,
force=args.force,
downscale_max_dim=args.downscale_max_dim,
downscale_suffix=args.downscale_suffix,
output_format=effective_output_format,
)
return i, None
except Exception as exc:
any_failed = True
print(f"{job_label} failed: {exc}", file=sys.stderr)
if args.fail_fast:
raise
return i, str(exc)
tasks = [asyncio.create_task(run_job(i, job)) for i, job in enumerate(jobs, start=1)]
try:
await asyncio.gather(*tasks)
except Exception:
for t in tasks:
if not t.done():
t.cancel()
raise
return 1 if any_failed else 0
def _generate_batch(args: argparse.Namespace) -> None:
exit_code = asyncio.run(_run_generate_batch(args))
if exit_code:
raise SystemExit(exit_code)
def _generate(args: argparse.Namespace) -> None:
prompt = _read_prompt(args.prompt, args.prompt_file)
prompt = _augment_prompt(args, prompt)
payload = {
"model": args.model,
"prompt": prompt,
"n": args.n,
"size": args.size,
"quality": args.quality,
"background": args.background,
"output_format": args.output_format,
"output_compression": args.output_compression,
"moderation": args.moderation,
}
payload = {k: v for k, v in payload.items() if v is not None}
output_format = _normalize_output_format(args.output_format)
_validate_transparency(args.background, output_format)
payload["output_format"] = output_format
output_paths = _build_output_paths(args.out, output_format, args.n, args.out_dir)
downscaled = None
if args.downscale_max_dim is not None:
downscaled = [str(_derive_downscale_path(p, args.downscale_suffix)) for p in output_paths]
if args.dry_run:
_print_request(
{
"endpoint": "/v1/images/generations",
"outputs": [str(p) for p in output_paths],
"outputs_downscaled": downscaled,
**payload,
}
)
return
print(
"Calling Image API (generation). This can take up to a couple of minutes.",
file=sys.stderr,
)
started = time.time()
client = _create_client()
result = client.images.generate(**payload)
elapsed = time.time() - started
print(f"Generation completed in {elapsed:.1f}s.", file=sys.stderr)
images = [item.b64_json for item in result.data]
_decode_write_and_downscale(
images,
output_paths,
force=args.force,
downscale_max_dim=args.downscale_max_dim,
downscale_suffix=args.downscale_suffix,
output_format=output_format,
)
def _edit(args: argparse.Namespace) -> None:
prompt = _read_prompt(args.prompt, args.prompt_file)
prompt = _augment_prompt(args, prompt)
image_paths = _check_image_paths(args.image)
mask_path = Path(args.mask) if args.mask else None
if mask_path:
if not mask_path.exists():
_die(f"Mask file not found: {mask_path}")
if mask_path.suffix.lower() != ".png":
_warn(f"Mask should be a PNG with an alpha channel: {mask_path}")
if mask_path.stat().st_size > MAX_IMAGE_BYTES:
_warn(f"Mask exceeds 50MB limit: {mask_path}")
payload = {
"model": args.model,
"prompt": prompt,
"n": args.n,
"size": args.size,
"quality": args.quality,
"background": args.background,
"output_format": args.output_format,
"output_compression": args.output_compression,
"input_fidelity": args.input_fidelity,
"moderation": args.moderation,
}
payload = {k: v for k, v in payload.items() if v is not None}
output_format = _normalize_output_format(args.output_format)
_validate_transparency(args.background, output_format)
payload["output_format"] = output_format
_validate_input_fidelity(args.input_fidelity)
output_paths = _build_output_paths(args.out, output_format, args.n, args.out_dir)
downscaled = None
if args.downscale_max_dim is not None:
downscaled = [str(_derive_downscale_path(p, args.downscale_suffix)) for p in output_paths]
if args.dry_run:
payload_preview = dict(payload)
payload_preview["image"] = [str(p) for p in image_paths]
if mask_path:
payload_preview["mask"] = str(mask_path)
_print_request(
{
"endpoint": "/v1/images/edits",
"outputs": [str(p) for p in output_paths],
"outputs_downscaled": downscaled,
**payload_preview,
}
)
return
print(
f"Calling Image API (edit) with {len(image_paths)} image(s).",
file=sys.stderr,
)
started = time.time()
client = _create_client()
with _open_files(image_paths) as image_files, _open_mask(mask_path) as mask_file:
request = dict(payload)
request["image"] = image_files if len(image_files) > 1 else image_files[0]
if mask_file is not None:
request["mask"] = mask_file
result = client.images.edit(**request)
elapsed = time.time() - started
print(f"Edit completed in {elapsed:.1f}s.", file=sys.stderr)
images = [item.b64_json for item in result.data]
_decode_write_and_downscale(
images,
output_paths,
force=args.force,
downscale_max_dim=args.downscale_max_dim,
downscale_suffix=args.downscale_suffix,
output_format=output_format,
)
def _open_files(paths: List[Path]):
return _FileBundle(paths)
def _open_mask(mask_path: Optional[Path]):
if mask_path is None:
return _NullContext()
return _SingleFile(mask_path)
class _NullContext:
def __enter__(self):
return None
def __exit__(self, exc_type, exc, tb):
return False
class _SingleFile:
def __init__(self, path: Path):
self._path = path
self._handle = None
def __enter__(self):
self._handle = self._path.open("rb")
return self._handle
def __exit__(self, exc_type, exc, tb):
if self._handle:
try:
self._handle.close()
except Exception:
pass
return False
class _FileBundle:
def __init__(self, paths: List[Path]):
self._paths = paths
self._handles: List[object] = []
def __enter__(self):
self._handles = [p.open("rb") for p in self._paths]
return self._handles
def __exit__(self, exc_type, exc, tb):
for handle in self._handles:
try:
handle.close()
except Exception:
pass
return False
def _add_shared_args(parser: argparse.ArgumentParser) -> None:
parser.add_argument("--model", default=DEFAULT_MODEL)
parser.add_argument("--prompt")
parser.add_argument("--prompt-file")
parser.add_argument("--n", type=int, default=1)
parser.add_argument("--size", default=DEFAULT_SIZE)
parser.add_argument("--quality", default=DEFAULT_QUALITY)
parser.add_argument("--background")
parser.add_argument("--output-format")
parser.add_argument("--output-compression", type=int)
parser.add_argument("--moderation")
parser.add_argument("--out", default=DEFAULT_OUTPUT_PATH)
parser.add_argument("--out-dir")
parser.add_argument("--force", action="store_true")
parser.add_argument("--dry-run", action="store_true")
parser.add_argument("--augment", dest="augment", action="store_true")
parser.add_argument("--no-augment", dest="augment", action="store_false")
parser.set_defaults(augment=True)
# Prompt augmentation hints
parser.add_argument("--use-case")
parser.add_argument("--scene")
parser.add_argument("--subject")
parser.add_argument("--style")
parser.add_argument("--composition")
parser.add_argument("--lighting")
parser.add_argument("--palette")
parser.add_argument("--materials")
parser.add_argument("--text")
parser.add_argument("--constraints")
parser.add_argument("--negative")
# Post-processing (optional): generate an additional downscaled copy for fast web loading.
parser.add_argument("--downscale-max-dim", type=int)
parser.add_argument("--downscale-suffix", default=DEFAULT_DOWNSCALE_SUFFIX)
def main() -> int:
parser = argparse.ArgumentParser(
description="Fallback CLI for explicit image generation or editing via GPT Image models"
)
subparsers = parser.add_subparsers(dest="command", required=True)
gen_parser = subparsers.add_parser("generate", help="Create a new image")
_add_shared_args(gen_parser)
gen_parser.set_defaults(func=_generate)
batch_parser = subparsers.add_parser(
"generate-batch",
help="Generate multiple prompts concurrently (JSONL input)",
)
_add_shared_args(batch_parser)
batch_parser.add_argument("--input", required=True, help="Path to JSONL file (one job per line)")
batch_parser.add_argument("--concurrency", type=int, default=DEFAULT_CONCURRENCY)
batch_parser.add_argument("--max-attempts", type=int, default=3)
batch_parser.add_argument("--fail-fast", action="store_true")
batch_parser.set_defaults(func=_generate_batch)
edit_parser = subparsers.add_parser("edit", help="Edit an existing image")
_add_shared_args(edit_parser)
edit_parser.add_argument("--image", action="append", required=True)
edit_parser.add_argument("--mask")
edit_parser.add_argument("--input-fidelity")
edit_parser.set_defaults(func=_edit)
args = parser.parse_args()
if args.n < 1 or args.n > 10:
_die("--n must be between 1 and 10")
if getattr(args, "concurrency", 1) < 1 or getattr(args, "concurrency", 1) > 25:
_die("--concurrency must be between 1 and 25")
if getattr(args, "max_attempts", 3) < 1 or getattr(args, "max_attempts", 3) > 10:
_die("--max-attempts must be between 1 and 10")
if args.output_compression is not None and not (0 <= args.output_compression <= 100):
_die("--output-compression must be between 0 and 100")
if args.command == "generate-batch" and not args.out_dir:
_die("generate-batch requires --out-dir")
if getattr(args, "downscale_max_dim", None) is not None and args.downscale_max_dim < 1:
_die("--downscale-max-dim must be >= 1")
_validate_model(args.model)
_validate_size(args.size, args.model)
_validate_quality(args.quality)
_validate_background(args.background)
_validate_model_specific_options(
model=args.model,
background=args.background,
input_fidelity=getattr(args, "input_fidelity", None),
)
_ensure_api_key(args.dry_run)
args.func(args)
return 0
if __name__ == "__main__":
raise SystemExit(main())

440
.codex-home/skills/.system/imagegen/scripts/remove_chroma_key.py Normal file
View file

@ -0,0 +1,440 @@
#!/usr/bin/env python3
"""Remove a solid chroma-key background from an image.
This helper supports the imagegen skill's built-in-first transparent workflow:
generate an image on a flat key color, then convert that key color to alpha.
"""
from __future__ import annotations
import argparse
from io import BytesIO
from pathlib import Path
import re
from statistics import median
import sys
from typing import Tuple
Color = Tuple[int, int, int]
KEY_DOMINANCE_THRESHOLD = 16.0
ALPHA_NOISE_FLOOR = 8
def _die(message: str, code: int = 1) -> None:
print(f"Error: {message}", file=sys.stderr)
raise SystemExit(code)
def _dependency_hint(package: str) -> str:
return (
"Activate the repo-selected environment first, then install it with "
f"`uv pip install {package}`. If this repo uses a local virtualenv, start with "
"`source .venv/bin/activate`; otherwise use this repo's configured shared fallback "
"environment."
)
def _load_pillow():
try:
from PIL import Image, ImageFilter
except ImportError:
_die(f"Pillow is required for chroma-key removal. {_dependency_hint('pillow')}")
return Image, ImageFilter
def _parse_key_color(raw: str) -> Color:
value = raw.strip()
match = re.fullmatch(r"#?([0-9a-fA-F]{6})", value)
if not match:
_die("key color must be a hex RGB value like #00ff00.")
hex_value = match.group(1)
return (
int(hex_value[0:2], 16),
int(hex_value[2:4], 16),
int(hex_value[4:6], 16),
)
def _validate_args(args: argparse.Namespace) -> None:
if args.tolerance < 0 or args.tolerance > 255:
_die("--tolerance must be between 0 and 255.")
if args.transparent_threshold < 0 or args.transparent_threshold > 255:
_die("--transparent-threshold must be between 0 and 255.")
if args.opaque_threshold < 0 or args.opaque_threshold > 255:
_die("--opaque-threshold must be between 0 and 255.")
if args.soft_matte and args.transparent_threshold >= args.opaque_threshold:
_die("--transparent-threshold must be lower than --opaque-threshold.")
if args.edge_feather < 0 or args.edge_feather > 64:
_die("--edge-feather must be between 0 and 64.")
if args.edge_contract < 0 or args.edge_contract > 16:
_die("--edge-contract must be between 0 and 16.")
src = Path(args.input)
if not src.exists():
_die(f"Input image not found: {src}")
out = Path(args.out)
if out.exists() and not args.force:
_die(f"Output already exists: {out} (use --force to overwrite)")
if out.suffix.lower() not in {".png", ".webp"}:
_die("--out must end in .png or .webp so the alpha channel is preserved.")
def _channel_distance(a: Color, b: Color) -> int:
return max(abs(a[0] - b[0]), abs(a[1] - b[1]), abs(a[2] - b[2]))
def _clamp_channel(value: float) -> int:
return max(0, min(255, int(round(value))))
def _smoothstep(value: float) -> float:
value = max(0.0, min(1.0, value))
return value * value * (3.0 - 2.0 * value)
def _soft_alpha(distance: int, transparent_threshold: float, opaque_threshold: float) -> int:
if distance <= transparent_threshold:
return 0
if distance >= opaque_threshold:
return 255
ratio = (float(distance) - transparent_threshold) / (
opaque_threshold - transparent_threshold
)
return _clamp_channel(255.0 * _smoothstep(ratio))
def _dominance_alpha(rgb: Color, key: Color) -> int:
spill_channels = _spill_channels(key)
if not spill_channels:
return 255
channels = [float(value) for value in rgb]
non_spill = [idx for idx in range(3) if idx not in spill_channels]
key_strength = (
min(channels[idx] for idx in spill_channels)
if len(spill_channels) > 1
else channels[spill_channels[0]]
)
non_key_strength = max((channels[idx] for idx in non_spill), default=0.0)
dominance = key_strength - non_key_strength
if dominance <= 0:
return 255
denominator = max(1.0, float(max(key)) - non_key_strength)
alpha = 1.0 - min(1.0, dominance / denominator)
return _clamp_channel(alpha * 255.0)
def _spill_channels(key: Color) -> list[int]:
key_max = max(key)
if key_max < 128:
return []
return [idx for idx, value in enumerate(key) if value >= key_max - 16 and value >= 128]
def _key_channel_dominance(rgb: Color, key: Color) -> float:
spill_channels = _spill_channels(key)
if not spill_channels:
return 0.0
channels = [float(value) for value in rgb]
non_spill = [idx for idx in range(3) if idx not in spill_channels]
key_strength = (
min(channels[idx] for idx in spill_channels)
if len(spill_channels) > 1
else channels[spill_channels[0]]
)
non_key_strength = max((channels[idx] for idx in non_spill), default=0.0)
return key_strength - non_key_strength
def _looks_key_colored(rgb: Color, key: Color, distance: int) -> bool:
if distance <= 32:
return True
spill_channels = _spill_channels(key)
if not spill_channels:
return True
return _key_channel_dominance(rgb, key) >= KEY_DOMINANCE_THRESHOLD
def _cleanup_spill(rgb: Color, key: Color, alpha: int = 255) -> Color:
if alpha >= 252:
return rgb
spill_channels = _spill_channels(key)
if not spill_channels:
return rgb
channels = [float(value) for value in rgb]
non_spill = [idx for idx in range(3) if idx not in spill_channels]
if non_spill:
anchor = max(channels[idx] for idx in non_spill)
cap = max(0.0, anchor - 1.0)
for idx in spill_channels:
if channels[idx] > cap:
channels[idx] = cap
return (
_clamp_channel(channels[0]),
_clamp_channel(channels[1]),
_clamp_channel(channels[2]),
)
def _apply_alpha_to_image(
image,
*,
key: Color,
tolerance: int,
spill_cleanup: bool,
soft_matte: bool,
transparent_threshold: float,
opaque_threshold: float,
) -> int:
pixels = image.load()
width, height = image.size
transparent = 0
for y in range(height):
for x in range(width):
red, green, blue, alpha = pixels[x, y]
rgb = (red, green, blue)
distance = _channel_distance(rgb, key)
key_like = _looks_key_colored(rgb, key, distance)
output_alpha = (
min(
_soft_alpha(distance, transparent_threshold, opaque_threshold),
_dominance_alpha(rgb, key),
)
if soft_matte and key_like
else (0 if distance <= tolerance else 255)
)
output_alpha = int(round(output_alpha * (alpha / 255.0)))
if 0 < output_alpha <= ALPHA_NOISE_FLOOR:
output_alpha = 0
if output_alpha == 0:
pixels[x, y] = (0, 0, 0, 0)
transparent += 1
continue
if spill_cleanup and key_like:
red, green, blue = _cleanup_spill(rgb, key, output_alpha)
pixels[x, y] = (red, green, blue, output_alpha)
return transparent
def _contract_alpha(image, pixels: int):
if pixels == 0:
return image
_, ImageFilter = _load_pillow()
alpha = image.getchannel("A")
for _ in range(pixels):
alpha = alpha.filter(ImageFilter.MinFilter(3))
image.putalpha(alpha)
return image
def _apply_edge_feather(image, radius: float):
if radius == 0:
return image
_, ImageFilter = _load_pillow()
alpha = image.getchannel("A")
alpha = alpha.filter(ImageFilter.GaussianBlur(radius=radius))
image.putalpha(alpha)
return image
def _encode_image(image, output_format: str) -> bytes:
out = BytesIO()
image.save(out, format=output_format.upper())
return out.getvalue()
def _alpha_counts(image) -> tuple[int, int, int]:
pixels = image.load()
width, height = image.size
total = 0
transparent = 0
partial = 0
for y in range(height):
for x in range(width):
alpha = pixels[x, y][3]
total += 1
if alpha == 0:
transparent += 1
elif alpha < 255:
partial += 1
return total, transparent, partial
def _sample_border_key(image, mode: str) -> Color:
width, height = image.size
pixels = image.load()
samples: list[Color] = []
if mode == "corners":
patch = max(1, min(width, height, 12))
boxes = [
(0, 0, patch, patch),
(width - patch, 0, width, patch),
(0, height - patch, patch, height),
(width - patch, height - patch, width, height),
]
for left, top, right, bottom in boxes:
for y in range(top, bottom):
for x in range(left, right):
red, green, blue = pixels[x, y][:3]
samples.append((red, green, blue))
else:
band = max(1, min(width, height, 6))
step = max(1, min(width, height) // 256)
for x in range(0, width, step):
for y in range(band):
red, green, blue = pixels[x, y][:3]
samples.append((red, green, blue))
red, green, blue = pixels[x, height - 1 - y][:3]
samples.append((red, green, blue))
for y in range(0, height, step):
for x in range(band):
red, green, blue = pixels[x, y][:3]
samples.append((red, green, blue))
red, green, blue = pixels[width - 1 - x, y][:3]
samples.append((red, green, blue))
if not samples:
_die("Could not sample background key color from image border.")
return (
int(round(median(sample[0] for sample in samples))),
int(round(median(sample[1] for sample in samples))),
int(round(median(sample[2] for sample in samples))),
)
def _remove_chroma_key(args: argparse.Namespace) -> None:
Image, _ = _load_pillow()
src = Path(args.input)
out = Path(args.out)
with Image.open(src) as image:
rgba = image.convert("RGBA")
key = (
_sample_border_key(rgba, args.auto_key)
if args.auto_key != "none"
else _parse_key_color(args.key_color)
)
transparent = _apply_alpha_to_image(
rgba,
key=key,
tolerance=args.tolerance,
spill_cleanup=args.spill_cleanup,
soft_matte=args.soft_matte,
transparent_threshold=args.transparent_threshold,
opaque_threshold=args.opaque_threshold,
)
rgba = _contract_alpha(rgba, args.edge_contract)
rgba = _apply_edge_feather(rgba, args.edge_feather)
total, transparent_after, partial_after = _alpha_counts(rgba)
out.parent.mkdir(parents=True, exist_ok=True)
output_format = "PNG" if out.suffix.lower() == ".png" else "WEBP"
out.write_bytes(_encode_image(rgba, output_format))
print(f"Wrote {out}")
print(f"Key color: #{key[0]:02x}{key[1]:02x}{key[2]:02x}")
print(f"Transparent pixels: {transparent_after}/{total}")
print(f"Partially transparent pixels: {partial_after}/{total}")
if transparent == 0:
print("Warning: no pixels matched the key color before feathering.", file=sys.stderr)
def _build_parser() -> argparse.ArgumentParser:
parser = argparse.ArgumentParser(
description="Remove a solid chroma-key background and write an image with alpha."
)
parser.add_argument("--input", required=True, help="Input image path.")
parser.add_argument("--out", required=True, help="Output .png or .webp path.")
parser.add_argument(
"--key-color",
default="#00ff00",
help="Hex RGB key color to remove, for example #00ff00.",
)
parser.add_argument(
"--tolerance",
type=int,
default=12,
help="Hard-key per-channel tolerance for matching the key color, 0-255.",
)
parser.add_argument(
"--auto-key",
choices=["none", "corners", "border"],
default="none",
help="Sample the key color from image corners or border instead of --key-color.",
)
parser.add_argument(
"--soft-matte",
action="store_true",
help="Use a smooth alpha ramp between transparent and opaque thresholds.",
)
parser.add_argument(
"--transparent-threshold",
type=float,
default=12.0,
help="Soft-matte distance at or below which pixels become fully transparent.",
)
parser.add_argument(
"--opaque-threshold",
type=float,
default=96.0,
help="Soft-matte distance at or above which pixels become fully opaque.",
)
parser.add_argument(
"--edge-feather",
type=float,
default=0.0,
help="Optional alpha blur radius for softened edges, 0-64.",
)
parser.add_argument(
"--edge-contract",
type=int,
default=0,
help="Shrink the visible alpha matte by this many pixels before feathering.",
)
parser.add_argument(
"--spill-cleanup",
dest="spill_cleanup",
action="store_true",
help="Reduce obvious key-color spill on opaque pixels.",
)
parser.add_argument(
"--despill",
dest="spill_cleanup",
action="store_true",
help="Alias for --spill-cleanup; decontaminate key-color edge spill.",
)
parser.add_argument("--force", action="store_true", help="Overwrite an existing output file.")
return parser
def main() -> None:
parser = _build_parser()
args = parser.parse_args()
_validate_args(args)
_remove_chroma_key(args)
if __name__ == "__main__":
main()

201
.codex-home/skills/.system/openai-docs/LICENSE.txt Normal file
View file

@ -0,0 +1,201 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf of
any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don\'t include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

38
.codex-home/skills/.system/openai-docs/SKILL.md Normal file
View file

@ -0,0 +1,38 @@
---
name: "openai-docs"
description: "Use for Codex models/pricing, scheduled tasks, skills, settings, setup, troubleshooting, customization, automations, and self-knowledge—including 'you,' 'your,' 'this app,' or 'this coding agent' when they refer to Codex—and for OpenAI APIs/products and ChatGPT Work. Also use for model choice/migration, prompting, SDKs, Responses, Realtime, agents, evals, and Chat/Work/Codex comparisons. Do not use for generic app/software tasks that merely mention Codex."
metadata:
short-description: "Codex models/pricing, scheduled tasks, skills, settings, setup, troubleshooting, and self-knowledge; OpenAI APIs and ChatGPT Work. 'You'/'this app' means Codex only."
---
# OpenAI Docs
Provide current, cited OpenAI product, API, model, and Codex guidance. Read zero or one primary reference.
**First substantive action:** Search the user's exact requested official OpenAI documentation topic and any explicitly named model using a concise, topic-specific query of 2-6 essential terms. When an already-available direct official documentation search and page-retrieval capability is present, use it first: search, then fetch or open the matching official page before general web search. Otherwise, immediately use official-domain web search, then actually open or fetch the relevant official page. Complete this source order before reading a reference, inspecting local or repository files, running a Codex manual or model resolver, drafting a plan, or answering from memory. Use the actual fetched page, not a search snippet or an unopened link. If one official search or page does not establish the answer, search another appropriate official domain and actually open or fetch the result. Preserve the exact requested model; never substitute a newer model.
**Only exception:** An explicitly requested, genuinely broad, cross-topic Codex setup, orientation, or system-map synthesis may use the manual first when shell execution and an allowed temporary cache are available. A specific Codex feature, setting, command, error, model, or requested citation remains docs-first. Mixed Chat/Work/Codex comparisons are official documentation questions, not manual-first Codex requests.
For generic software tasks, answer the software task directly. OpenAI implementation, debugging, SDK, API, prompting, agent, and eval requests are not generic.
For a straightforward factual or citation-only request, follow the source order and do not read a route reference. This includes straightforward API facts, ChatGPT Work or mixed Chat/Work/Codex comparisons, model tiers, aliases, Pro mode, reasoning settings, factual migration baselines, and narrow Codex facts. Prioritize `learn.chatgpt.com` for ChatGPT Work.
## Choose one primary route
Use the first matching route, and read its reference only when the requested task needs that specialized workflow:
- **Explicitly requested local documentation integration:** Read [integration guidance](references/mcp-diagnostics.md) only when the user explicitly requests that local integration.
- **Model migration, upgrades, or model-specific prompting:** Read [model-migration.md](references/model-migration.md) for actual migration planning, implementation, dynamic target resolution, or prompt changes. Preserve an explicitly requested target.
- **Model selection and comparisons:** Read [model-selection.md](references/model-selection.md) only when nuanced current, latest, default, cost, latency, quality, or modality tradeoffs need more guidance. Do not run a migration resolver for selection alone.
- **Product, API, ChatGPT Work, and mixed Chat/Work/Codex documentation:** Read [official-docs.md](references/official-docs.md) only when fetched official pages leave source selection, API schemas, or the requested implementation unresolved. This route is not manual-first.
- **Explicitly broad Codex setup, orientation, or cross-topic synthesis:** Read [codex-self-knowledge.md](references/codex-self-knowledge.md) when the eligible Codex manual or deeper Codex procedures are needed.
Read at most one primary reference. Do not open every route, bundled model guide, or helper script. Read a supporting reference or run a helper only when the chosen workflow demonstrably needs it.
## Source and execution boundaries
- Search, open, fetch, and cite only `developers.openai.com`, `platform.openai.com`, and `learn.chatgpt.com`. Cite the page that supports the claim. State uncertainty when official sources do not establish pricing, availability, account access, limits, or behavior.
- Preserve an explicitly requested model for selection, migration, and prompting. Resolve an unspecified latest or current migration target only after searching and fetching current official guidance.
- Use `references/latest-model.md` only as a disclosed fallback after current official model guidance does not answer the question. Read `references/upgrading-to-gpt-5p6-sol.md` only for an actual, requested GPT-5.6-family migration; read `references/prompting-guide.md` only for requested prompting work.
- Before building, running, editing, debugging, or testing an API-backed app or tool, use `openai-platform-api-key` first when available. Documentation, conceptual examples, model selection, and read-only guidance do not require an API key.
- Say "OpenAI Docs" or "official OpenAI documentation" in user-facing answers. Keep exact official citations and examples concise.

6
.codex-home/skills/.system/openai-docs/agents/openai.yaml Normal file
View file

@ -0,0 +1,6 @@
interface:
display_name: "OpenAI Docs"
short_description: "OpenAI and Codex docs for models, skills, tasks, and setup"
icon_small: "./assets/openai-small.svg"
icon_large: "./assets/openai.png"
default_prompt: "Use OpenAI Docs for official docs lookup, questions about Codex itself or Codex surfaces, model selection, model migration, and prompt-upgrade work."

3
.codex-home/skills/.system/openai-docs/assets/openai-small.svg Normal file
View file

@ -0,0 +1,3 @@
<svg xmlns="http://www.w3.org/2000/svg" width="14" height="14" fill="currentColor" viewBox="0 0 14 14">
<path d="M10.931 3.34a.112.112 0 0 0-.069-.104l-.038-.007c-1.537.05-2.45.318-3.714 1.002v6.683c.48-.248.936-.44 1.414-.58.695-.203 1.417-.292 2.303-.305l.038-.008a.113.113 0 0 0 .066-.104V3.341ZM2.363 9.919c0 .064.051.11.105.111l.33.008c1.162.046 2.042.243 2.975.662-.403-.585-1.008-1.075-1.654-1.292a.991.991 0 0 1-.674-.941v-5.14a6.36 6.36 0 0 0-.59-.076l-.37-.02a.115.115 0 0 0-.122.111v6.577Zm9.455-.001a.998.998 0 0 1-.877.992l-.101.007c-.832.012-1.47.095-2.066.27-.599.174-1.176.448-1.883.863a.444.444 0 0 1-.449 0c-1.299-.763-2.229-1.07-3.689-1.125l-.299-.008a.997.997 0 0 1-.977-.998V3.342c0-.573.478-1.017 1.038-.999l.417.023c.188.015.35.037.513.062v-.754c0-.708.749-1.244 1.429-.903.984.492 1.836 1.449 2.15 2.505 1.216-.617 2.222-.884 3.771-.934l.105.003a.998.998 0 0 1 .918.996v6.576ZM4.332 8.466c0 .049.03.087.07.1l.24.091a4.319 4.319 0 0 1 1.581 1.176V3.721c-.164-.803-.799-1.617-1.584-2.07l-.162-.088c-.025-.012-.054-.013-.088.009a.12.12 0 0 0-.057.102v6.792Z"/>
</svg>
Side by side

After

Width:  |  Height:  |  Size: 1.1 KiB

BIN
.codex-home/skills/.system/openai-docs/assets/openai.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 1.4 KiB

71
.codex-home/skills/.system/openai-docs/references/codex-self-knowledge.md Normal file
View file

@ -0,0 +1,71 @@
# Codex self-knowledge
Use this manual-first route only for genuinely broad Codex setup, orientation, customization, troubleshooting, local-state guidance, or system-map synthesis across skills, plugins, MCP, hooks, `AGENTS.md`, automations, and product surfaces. Mixed Chat/Work/Codex comparisons belong to `official-docs.md` instead.
Narrow Codex documentation questions require official documentation search first, then an actual page open or fetch using an available documentation or official-domain web capability. This includes a single feature such as Codex Goals, a specific setting, documented behavior, exact error, or requested page citation. Search and fetch the exact official topic before inspecting local files or bundled references. Do not fetch the manual, read bundled references, inspect local configuration or caches, or turn a targeted documentation lookup into broad product synthesis. Current or latest model questions follow the model-selection route.
## Start with the manual
Reuse a manual path and outline path already established in the same thread when both remain usable and current. Refresh before relying on a path fetched more than about a day ago, obtained from another thread or uncertain source, or missing likely-current information.
Otherwise, run the bundled manual helper first. Skip it without probing only when policy explicitly makes the session read-only, shell execution unavailable, or every allowed temporary cache location unavailable. Workspace-only write access is not enough: the helper needs an allowed writable temp cache. A guessed sandbox restriction is not evidence that the helper is unavailable.
Resolve `<skill-dir>` to the actual installed skill directory, then run:
```bash
node <skill-dir>/scripts/fetch-codex-manual.mjs
```
The helper automatically chooses the first usable cache location in this order:
1. `$TMPDIR/openai-docs-cache`
2. `%TEMP%\openai-docs-cache`
3. `%TMP%\openai-docs-cache`
4. `/private/tmp/openai-docs-cache`
5. `/tmp/openai-docs-cache`
Use an explicit override only when the allowed cache must be selected manually:
```bash
node <skill-dir>/scripts/fetch-codex-manual.mjs --cache-dir <cache-dir>
```
On Windows, `%TEMP%` and `%TMP%` are discovered automatically; `$env:TEMP\openai-docs-cache` is a typical PowerShell override. The helper handles configured HTTP(S) proxies and falls back to `curl` when needed. Do not require a POSIX-only environment prefix or an unnecessary cache override.
The helper verifies the current source and returns a manual path, outline path, freshness status, and heading outline. Use that outline to locate relevant headings and line ranges, then read or search only the returned manual and outline paths. Do not inspect unrelated repositories, caches, source trees, or local state to establish a public Codex product claim.
For follow-up questions in the same thread, reuse those fresh paths instead of fetching again. If asked whether the manual is current enough to rely on now, rerun the helper when an allowed temp cache is available and answer from its reported status and returned paths.
## Fill only genuine documentation gaps
If the manual answers a claim, stop retrieving sources for that claim. Its official source pages and known anchors are sufficient citation support. Continue the user's broader task when the documentation lookup was only one dependency.
If the helper was legitimately skipped, actually fails, or the fresh manual lacks a material or likely-current claim, use the narrowest official follow-up. Search the exact topic using an available documentation or approved-domain web capability, then actually open or fetch a clearly relevant official result. A page-specific citation can justify the same narrow follow-up.
For an undocumented Codex term, mode, acronym, or exact error, first check adjacent manual terminology. Map it to the closest documented concept when possible. If the exact term is material or likely current, perform one targeted official search-and-fetch; if it remains undocumented, say so. Do not expand into internal knowledge bases, private source trees, guessed roadmap details, or account-specific workarounds.
If official documentation conflicts with a callable capability verified in the current session, explicitly state the conflict and prefer that verified behavior for this environment. Otherwise, resolve unsupported claims with bounded uncertainty or route the user to support, an administrator, or product feedback.
## Choose the smallest matching Codex surface
- Prompt or thread context: one-off task constraints.
- Repository `AGENTS.md`: durable team conventions, commands, and verification expectations; nested files apply more specifically within their subtree.
- Project `.codex/config.toml`: settings for a trusted repository, including sandbox, MCP, hooks, model, and reasoning defaults.
- Global config or global guidance: personal defaults across repositories.
- Skill: a reusable workflow, optionally with focused references or scripts.
- Plugin: an installable bundle of skills, tools, commands, MCP configuration, hooks, apps, assets, or related metadata.
- MCP server or app connector: authorized live external data and actions. Use an authenticated connector, not web search or memory, for private Google Docs, Calendar, Slack, GitHub, Notion, or similar workspace data.
- Automation: scheduled checks, reminders, monitors, or follow-ups; use an existing-thread heartbeat when continuity matters.
- Hook: mechanical enforcement around lifecycle events, tool calls, commands, or edits.
Split requests that combine one-off, durable, repository-scoped, and recurring behavior instead of forcing them onto a single surface. For example, "always do this, but only for this PR" belongs in the current prompt or thread unless the user explicitly wants persistence or enforcement.
For a surface recommendation, state what to use, why it fits, what to avoid, and the manual or official documentation supporting the answer.
For product surfaces, distinguish terminal-first CLI work, editor-attached IDE work, desktop planning or review, hosted cloud execution, in-app browser testing, the user's existing Chrome session, and desktop Computer Use. Keep `config.toml` defaults, `requirements.toml` constraints, and managed or administrator policy separate. An API key does not establish ChatGPT, Codex cloud, connector, or account access.
For plugin or app failures, check the installed bundle, enabled state, connector authorization, MCP setup, restart or new-thread expectations, and workspace policy before inferring a cause. Route billing, entitlements, undocumented rollout labels, and unsupported access paths to the appropriate support or administrative owner.
Memory can provide user preferences or context, but explicit prompt instructions win and memory is not a source for current external facts. Sandbox or network denials require narrowly scoped escalation with a clear justification; destructive commands, writes outside the workspace, and broad access changes require explicit approval.
When a page-specific citation helps, useful official anchors include `concepts/customization#agents-guidance`, `concepts/customization#skills`, `plugins/build#plugin-structure`, `concepts/customization#mcp`, `config-advanced#hooks`, `app/automations#thread-automations`, and `config-reference#configtoml`.

25
.codex-home/skills/.system/openai-docs/references/latest-model.md Normal file
View file

@ -0,0 +1,25 @@
# Latest model fallback
This is a compact, non-authoritative fallback, not a source for current availability, prices, aliases, or defaults. First search for and fetch current official model guidance at `https://developers.openai.com/api/docs/guides/latest-model` and the relevant official model page. The fetched official documentation wins if this snapshot has drifted. Disclose any use of this fallback.
## GPT-5.6 family
| Model ID | Documented workload to verify against the current model page |
| --- | --- |
| `gpt-5.6` | GPT-5.6 family alias; verify its currently documented routing and availability. |
| `gpt-5.6-sol` | Quality-first flagship, reasoning, and difficult coding work. |
| `gpt-5.6-terra` | Balanced quality, latency, and cost. |
| `gpt-5.6-luna` | High-throughput, lower-latency work. |
Use `https://developers.openai.com/api/docs/guides/upgrading-to-gpt-5p6-sol` for an actual GPT-5.6 migration and `https://developers.openai.com/api/docs/guides/prompt-guidance-gpt-5p6` for requested GPT-5.6 prompting. Open and read the relevant page before recommending a request shape, reasoning setting, endpoint, tool behavior, or migration.
## Explicitly requested existing models
| Model ID | Boundary |
| --- | --- |
| `gpt-4.1` | Preserve only when the user explicitly requests this model or existing migration target; search and fetch its own current official guide. |
| `gpt-5.4` | Preserve only when the user explicitly requests this model or existing migration target; search and fetch its own current official guide. |
Do not promote a legacy model as the current default, substitute it into an unrelated task, or replace an explicitly requested legacy target with GPT-5.6. Recommend a specialized image, audio, realtime, coding, moderation, or embedding model only after verifying the requested modality against current official documentation.
Verify GPT-5.6 Pro against current official Responses and model documentation before describing model IDs, reasoning modes, request parameters, or account availability; do not invent a separate `gpt-5.6-pro` model slug.

27
.codex-home/skills/.system/openai-docs/references/mcp-diagnostics.md Normal file
View file

@ -0,0 +1,27 @@
# Local documentation MCP setup and diagnostics
Use this route only when the user explicitly asks to configure or troubleshoot the official OpenAI documentation MCP server in a supported **local Codex client**. A missing documentation tool during an ordinary documentation request is not a setup request: answer with the root skill's official-domain web fallback without installation, sandbox escalation, configuration changes, or restart.
## Verify the supported local setup
1. Search and fetch current official Codex MCP setup documentation when those tools are callable. Otherwise, search and fetch the relevant official OpenAI documentation directly.
2. Confirm the documented endpoint is `https://developers.openai.com/mcp` and verify the supported command or configuration against that current documentation before recommending it.
3. When the current documentation supports it, the local Codex CLI setup is:
```sh
codex mcp add openaiDeveloperDocs --url https://developers.openai.com/mcp
```
The equivalent documented configuration is:
```toml
[mcp_servers.openaiDeveloperDocs]
url = "https://developers.openai.com/mcp"
```
4. Check the supported local client's MCP listing or configuration, its enabled state, relevant workspace/admin policy, and any documented authentication requirement. Verify success from the actual command result, configuration, or a callable documentation-tool search/fetch; never claim installation or access without evidence.
5. Recommend a local-client restart or new local session only when current official documentation or observed client behavior requires it. Clearly identify which local client must refresh.
A skill dependency declaration, configured server, or local-client setup does not make a tool callable in an already running session. In particular, editing a hosted container's local configuration cannot install a tool into the host or model's current tool inventory. Never claim a local command installed the server into a current hosted session.
Only perform a local installation or configuration change when the user explicitly authorizes that change. Never request sandbox escalation, edit hosted configuration, install a dependency, or ask the user to restart a hosted session merely to answer an ordinary documentation question.

45
.codex-home/skills/.system/openai-docs/references/model-migration.md Normal file
View file

@ -0,0 +1,45 @@
# Model migration and prompting
Use this route for model upgrades, migration planning, model-specific prompting, or latest/current/default prompting guidance. First search current official OpenAI documentation for the exact requested topic and model, then open or fetch the relevant official page using an available documentation or official-domain web capability. Do not run a resolver, open bundled references, or rely on a guide URL before completing that official search and actual page fetch.
## Choose the target before loading more context
- **Explicit model target:** Preserve the user's exact requested target, including an explicitly requested GPT-4.1 or GPT-5.4 migration. Do not run the latest-model resolver and do not substitute a newer model. Search for and fetch current guidance for that exact model. A GPT-5.4 migration must not load GPT-5.6 guidance or references.
- **Unspecified, latest, current, or default target:** Search for and fetch `https://developers.openai.com/api/docs/guides/latest-model` first. Use the corresponding `latest-model.md` metadata only when dynamic migration resolution is needed, then run the platform-specific resolver below and preserve its returned model and exact guide URLs.
- **Latest/current/default prompting:** Follow the dynamic-target route, then use the returned prompting guide. Do not run the resolver for explicitly named-model prompting.
- **Pure model selection:** Use `references/model-selection.md` instead. Do not run the resolver.
For POSIX shells, invoke the resolver through `sh`, without assuming an executable bit:
```sh
sh <skill-dir>/scripts/resolve-latest-model-info
```
On Windows, use the CommonJS entry point with Node.js 18 or newer:
```text
node <skill-dir>\scripts\resolve-latest-model-info.cjs
```
If the Windows Node runtime is unavailable and `load_workspace_dependencies` is callable, use its returned runtime and retry once. Do not execute the extensionless POSIX wrapper directly on Windows.
Do not suppress or redirect resolver stdout. Success requires JSON with nonempty `model`, `migrationGuideUrl`, and `promptingGuideUrl` fields. If the command fails or any required field is missing, retry the platform-specific command once, then fall back to current official documentation and finally disclosed bundled references.
## Retrieve only the guidance this request needs
Treat returned guide URLs as opaque: fetch those exact URLs without deriving, substituting, or appending a model query. Use an available official documentation or first-party-domain capability to open and read the relevant official page. Retry the exact guide URL when its response contains only a title or no substantive body.
- Fetch `migrationGuideUrl` for a requested migration or upgrade plan.
- Fetch `promptingGuideUrl` only when the user asks for prompting guidance or the migration requires a prompt change. Extract only `## Prompting Best Practices` through the next H2 heading.
- For explicitly named-model prompting, fetch that model's official prompting guidance and extract only `## Prompting Best Practices` through the next H2 heading. Do not load a migration reference or run the resolver.
- For an actual GPT-5.6-family migration or implementation plan, fetch `https://developers.openai.com/api/docs/guides/upgrading-to-gpt-5p6-sol`. For specifically requested GPT-5.6 prompting, fetch `https://developers.openai.com/api/docs/guides/prompt-guidance-gpt-5p6`. Read `references/upgrading-to-gpt-5p6-sol.md` only when fetched official guidance does not resolve needed compatibility gates, scoped code changes, tier-aware routing, validation, or other migration-specific judgment. Never load it for documentation-only questions about model tiers, the family alias, Pro mode, reasoning effort, or current guidance when the fetched official documentation already answers them.
- Read `references/prompting-guide.md` only when prompting guidance or prompt changes are actually needed and current official guidance is unavailable.
- Read `references/upgrade-guide.md` only when current official migration guidance is unavailable. Disclose when a bundled fallback was used.
## Keep implementation changes scoped
Change active model defaults and directly related prompt surfaces only when the user requested that work. Update registries, model pickers, capability metadata, routing, pricing, or tests only when they are in scope and current official documentation verifies the relevant values.
Preserve each workload's cost, latency, quality, reasoning, tool, endpoint, and output-contract role. Do not collapse a tiered router into one flagship model, replace intentionally pinned fallbacks, or rewrite historical examples, fixtures, eval baselines, provider comparisons, or unrelated SDK and authentication configuration.
If a safe migration requires an endpoint change, request-schema change, tool-handler change, or other implementation outside the requested scope, report the exact compatibility blocker and smallest follow-up instead of silently changing behavior.

12
.codex-home/skills/.system/openai-docs/references/model-selection.md Normal file
View file

@ -0,0 +1,12 @@
# Model selection
Use this route for model recommendations, comparisons, and latest/current/default choices when the user is not requesting a migration or prompting guidance.
1. Search current official OpenAI documentation for the exact requested workload and any explicitly named model; then open or fetch the relevant official page. For current or latest family guidance, use `https://developers.openai.com/api/docs/guides/latest-model`.
2. Use any available official documentation or first-party-domain search. Read the actual source; do not make a recommendation from a search snippet, guessed default, or bundled snapshot.
3. Match the documented model to the user's requested modality, quality, latency, cost, context, and workload. Distinguish flagship, balanced, high-throughput, coding, audio, image, or other specialized roles only when the fetched current documentation supports the distinction.
4. Preserve an explicitly requested model or existing target. Cite the current official page and state uncertainty about availability, pricing, limits, or account access.
Pure model selection does not require migration metadata. **Do not run the resolver.**
Read `references/latest-model.md` only when fetched current official sources cannot answer the question. Disclose that bundled fallback guidance was used and may be outdated.

25
.codex-home/skills/.system/openai-docs/references/official-docs.md Normal file
View file

@ -0,0 +1,25 @@
# Official documentation, API references, and ChatGPT Work
Use this route for OpenAI product or API documentation, examples, citations, ChatGPT Work, learning content, mixed Chat/Work/Codex comparisons, and narrow Codex product documentation. Follow the root skill's official-source order and credential boundary.
An explicit OpenAI documentation question stays documentation-first even when embedded in a broader repository, Promptfoo, agent-evaluation, `PLANS.md`, frontend, tool-use, image, Realtime API, SDK installation, or streaming-debugging task. Search the exact requested official documentation and open or fetch its relevant page before inspecting local files, drafting a plan, running evals, reading bundled references, or invoking the Codex manual. Then use the fetched official source to complete the requested work.
## Find the smallest useful source
1. Search the exact topic with a specific, title-like query containing 2-6 essential terms. Prefer an already-available direct official documentation search and page-retrieval capability; search, then fetch or open the best page or section. Otherwise, immediately use official-domain web search and actually open or fetch the result.
2. If the results are noisy, narrow the query. When a plausible official documentation URL is available, open or fetch the page instead of relying on search snippets. Use an already available documentation index only when there is no clear search query.
3. For API schemas, required fields, parameters, or endpoint shapes, use an already available OpenAPI or specification capability when it directly resolves the question. Otherwise verify the shape against the fetched official API guide or reference.
4. Cite the exact official page supporting each consequential claim. Keep examples minimal, paraphrase instead of quoting at length, and state when official sources do not establish a capability, parameter, price, or availability.
Preserve an explicitly requested model in the search and answer. Search the requested topic directly for model-specific frontend and tool-use prompting, image input or generation, Realtime voice or translation, official Agents SDK installation, Responses streaming errors, and Codex Goals. A specific Codex feature, error, setting, or requested citation is a narrow documentation lookup, not the broad manual-first exception. Current or latest model recommendations follow the model-selection route and the same official search-and-fetch order.
## ChatGPT Work and mixed surfaces
Treat a comparison between Chat, Work, and Codex as ChatGPT Work documentation, not broad Codex self-knowledge. Search `learn.chatgpt.com` and open or fetch the relevant official page. Useful starting pages are:
- `https://learn.chatgpt.com/docs/use-chatgpt`
- `https://learn.chatgpt.com/docs/get-started-with-work`
If someone simply asks an API question from ChatGPT Work, answer the API question from the relevant API guide. Being in Work does not make it a question about the Work product.
Separate documented user-facing purposes from unsupported claims about underlying models, hard capability boundaries, file or context inheritance, exact UI labels, account entitlements, and rollout availability. When those details cannot be verified, cite the closest allowed official source and state the uncertainty.

287
.codex-home/skills/.system/openai-docs/references/prompting-guide.md Normal file
View file

@ -0,0 +1,287 @@
## Retrieve the live GPT-5.6 prompting guidance
Use already-callable official documentation search and fetch, or immediately use official-domain web search and fetch, to retrieve the live GPT-5.6 prompting guidance from:
https://developers.openai.com/api/docs/guides/model-guidance?model=gpt-5.6#prompting-best-practices
Read only the `## Prompting Best Practices` section, stopping at the next H2 heading. The URL anchor points to the section visually, but a documentation fetch may return the full page, so explicitly extract only that section.
Treat the live section as the canonical model-specific prompting guidance. Use the local guidance below only for skill-specific migration judgment: deciding what to preserve, remove, rewrite, or test when adapting an existing prompt stack to GPT-5.6.
## Skill-specific migration judgment
GPT-5.6 works best when prompts define the outcome, important constraints, available evidence, and completion bar, then leave room for the model to choose an efficient path. Compared with earlier GPT-5 models, many applications can use shorter prompts and smaller tool sets without losing quality.
Do not carry over every instruction from an older prompt stack. Legacy prompts often repeat rules, prescribe unnecessary steps, expose irrelevant tools, or include examples that no longer change behavior. With GPT-5.6, this can encourage extra exploration, repeated validation, and larger accumulated context.
Start with the smallest prompt and tool set that passes your evals. Add an instruction, example, or tool only when it fixes a measured failure mode.
## Simplify prompts first
When migrating an existing prompt, remove redundant scaffolding before adding new GPT-5.6-specific instructions.
Trim:
- repeated statements of the same rule;
- generic “be thorough,” “be concise,” or “think step by step” language;
- examples that do not change behavior;
- process instructions for behavior the model already performs reliably;
- tools and tool descriptions unrelated to the task.
Keep:
- the user-visible outcome;
- success criteria and stopping conditions;
- safety, business, evidence, and permission constraints;
- tool-routing rules when the correct route is not obvious;
- required output shape and validation requirements.
Review the remaining instructions for contradictions. GPT-5-class models follow prompt contracts closely, so conflicting rules can create more instability than missing detail.
## Outcome-first prompts and stopping conditions
Describe the destination rather than prescribing every step. GPT-5.6 can usually choose an efficient search, tool, or reasoning path when the prompt states what good looks like.
Prefer:
Resolve the customer's issue end to end.
Success means:
- make the eligibility decision from available policy and account evidence
- complete any allowed action before responding
- return completed_actions, customer_message, and blockers
- if required evidence is missing, ask for the smallest missing field
Avoid unnecessary absolute rules. Use ALWAYS, NEVER, must, and only for true invariants such as safety rules, required fields, or actions that should never happen. For judgment calls, such as when to search, ask, use a tool, or keep iterating, prefer decision rules.
Preserve explicit user values. When the correct value is implicit, provide decision criteria and let the model reason from context or schema. Avoid universal defaults, keyword maps, and broad semantic shortcuts.
Add stopping conditions:
Resolve the request in the fewest useful tool loops, but do not let loop
minimization outrank correctness, required evidence, calculations, or
required citations.
After each result, ask whether the core request can now be answered with
useful evidence. If yes, answer. If required evidence is still missing,
name the missing fact and use the smallest useful fallback.
## Personality, collaboration, and response length
GPT-5.6 is efficient, direct, and more compressed than recent models. For customer-facing assistants and collaborative products, define both personality and collaboration style.
- Personality controls tone, warmth, directness, formality, humor, empathy, and polish.
- Collaboration style controls when the model asks questions, makes assumptions, takes initiative, explains tradeoffs, checks work, and handles uncertainty.
Keep both short. Personality should shape the user experience; collaboration instructions should shape task behavior. Neither should replace clear goals, success criteria, tool rules, or stopping conditions.
Use concrete writing controls:
Lead with the conclusion. Include the evidence needed to support it, any
material caveat, and the next action. Keep all required facts, decisions,
caveats, and next steps. Trim introductions, repetition, generic reassurance,
and optional background first.
Avoid generic “be brief,” “keep it short,” or “use minimal text” instructions. GPT-5.6 is already biased toward compression, and generic brevity can make it omit required evidence or parts of an artifact.
For customer-facing tone, prefer concrete guidance:
Be direct and tactful. Acknowledge friction specifically when relevant.
Avoid canned reassurance and unnecessary sign-offs.
Avoid blanket language rules such as “always respond in the user's language” unless that is truly the product requirement. Specify the intended output language and when it should change.
For editing, rewriting, summaries, and customer-facing drafts, tell the model what to preserve:
Preserve the requested artifact, length, structure, genre, and factual claims
first. Improve clarity, flow, and correctness without adding new claims,
sections, or a more promotional tone unless requested.
## Autonomy and permissions
GPT-5.6 can be proactive and persistent. Define which level of action each request authorizes.
For requests to answer, explain, review, diagnose, or plan, inspect the
relevant materials and report the result. Do not implement changes unless
the request also asks for them.
For requests to change, build, or fix, make the requested in-scope local
changes and run relevant non-destructive validation without asking first.
Require confirmation for external writes, destructive actions, purchases,
or a material expansion of scope.
Specify which local actions are safe without approval, such as reading files, inspecting logs, searching, editing in-scope code, and running non-destructive tests.
Avoid repeating “ask first” throughout the prompt. Repetition can cause unnecessary permission checks even for safe, expected actions.
For long-running work, define the current layer of work. Distinguish research, design, implementation, review, and external coordination so the model does not silently move from one layer to another.
## Tool routing
Expose only task-relevant tools. Tool descriptions should state what the tool does, when to use it, important return fields, and error behavior.
When correctness depends on prerequisite retrieval or lookup, say so:
Before taking an action, resolve required discovery, retrieval, and
validation steps. Do not skip a prerequisite because the intended final
state seems obvious.
When several reads are independent, parallelize them. When one result determines the next action, keep the work sequential. After parallel retrieval, synthesize before acting.
If a tool returns empty, partial, or suspiciously narrow results, try one or two meaningful fallbacks before concluding that no result exists.
## Programmatic Tool Calling
Programmatic Tool Calling is useful when code can reduce large, structured intermediate results before they return to model context.
Use it for:
- filtering, joining, sorting, ranking, deduplication, and aggregation;
- batching across many similar records;
- repeated deterministic validation;
- large structured results that can be reduced to a compact schema.
Prefer direct tool calls when:
- one call is sufficient;
- intermediate outputs are already small;
- each result may change the next decision;
- an action requires approval;
- the final answer must preserve citations or native artifacts;
- the workflow requires semantic judgment between calls.
Do not rely on generic instructions such as “use Programmatic Tool Calling efficiently.” State the bounded stage, eligible tools, output schema, retry limit, stop condition, and handoff back to direct model judgment.
Use Programmatic Tool Calling only for the bounded record-reduction stage.
Call only the documented read-only tools. Filter and deduplicate the
intermediate results, then emit exactly the required compact schema with
evidence fields. Retry transient failures at most twice. Use direct tool
calls for approval, semantic judgment, citations, and final validation.
Evaluate the final user-visible answer, not only the program result. Lower tokens, latency, calls, or turns are improvements only when the final answer still meets the required quality bar.
## Grounding, citations, and retrieval budgets
For grounded answers, citation behavior should be part of the prompt. Define what needs support, what counts as enough evidence, and how to behave when evidence is missing. Absence of evidence should not automatically become a factual “no.”
For ordinary Q&A, start with one broad search using short, discriminative
keywords. If the top results contain enough support for the core request,
answer from those results.
Make another retrieval call only when a required fact, owner, date, ID, or
source is missing; the user asked for exhaustive coverage or comparison; a
specific artifact must be read; or an important claim would otherwise be
unsupported.
Do not search again only to improve phrasing, add examples, or support
nonessential detail.
For research and synthesis:
- cite only retrieved sources;
- attach citations to the claims they support;
- label inference separately from directly supported facts;
- state conflicts between sources;
- narrow the answer or report missing evidence instead of guessing.
For creative drafting, distinguish source-backed facts from creative wording. Do not invent names, metrics, dates, roadmap status, customer outcomes, or product capabilities to make a draft sound stronger.
## Long-running workflows and state
For multi-step or tool-heavy tasks, prompt for a short visible preamble before the first tool call, then sparse outcome-based updates at major phase changes. Do not ask the model to narrate routine tool calls.
Before tool calls for a multi-step task, send a one- or two-sentence
user-visible update that states the first step. During the task, update only
when a major phase begins or a finding changes the plan. Each update should
state one concrete outcome and the next step.
Preserve assistant phase values when replaying history so the model can distinguish commentary from the final answer. If using previous_response_id, prior assistant state is preserved automatically. If replaying history manually, preserve each original phase value unchanged.
Compact after major milestones rather than every turn. Keep the prompt functionally consistent after compaction and treat compacted items as opaque state.
Persisted reasoning is useful when the objective, assumptions, and priorities remain stable across turns. Use current-turn behavior when earlier reasoning is no longer relevant. Do not treat persisted reasoning as an always-on optimization: stale reasoning can add tokens, increase latency, and anchor the model to an outdated approach.
Prompt caching also affects prompt construction. Keep reusable prefixes stable and avoid unnecessary churn in large system prompts. Use explicit cache breakpoints only when they improve measured cache behavior and cost for the workload.
## Reasoning effort
Treat reasoning effort as a last-mile tuning knob, not the first response to a weak result.
- Preserve the current GPT-5.5 or GPT-5.4 reasoning effort as the baseline.
- Test the same setting and one level lower on representative tasks.
- Use low for latency-sensitive work when it preserves quality.
- Use medium as a balanced starting point.
- Use high or xhigh only when evals show a meaningful gain.
- Reserve max for the hardest quality-first workloads; do not recommend it globally.
Before increasing reasoning effort, check whether the prompt is missing a success criterion, dependency rule, tool-routing rule, or verification loop.
## Frontend and visual tasks
GPT-5.6 has stronger layout, visual hierarchy, and design judgment. Still provide product context, preserve the existing design system, and name the states and constraints that matter.
For incremental frontend changes:
- inspect and preserve existing design tokens, components, and patterns;
- do not add extra features or decorative UI unless requested;
- preserve responsive behavior and expected states;
- render and inspect the result before finalizing.
For vision, computer use, localization, or OCR tasks where spatial precision matters, choose image detail intentionally. Use original detail for large, dense, or coordinate-sensitive images when the extra input cost and latency are justified.
## Check work before finishing
Give GPT-5.6 access to tools that can validate the output, and state what validation matters.
For coding:
After making changes, run the most relevant validation available:
- targeted tests for changed behavior
- type checks or lint checks when applicable
- build checks for affected packages
- a minimal smoke test when full validation is too expensive
If validation cannot be run, explain why and describe the next best check.
For visual artifacts:
Render the artifact before finalizing. Inspect layout, clipping, spacing,
missing content, and visual consistency. Revise until the rendered output
matches the requirements.
For implementation plans, include requirements, named resources or files, state transitions or data flow, validation checks, failure behavior, privacy or security considerations, and open questions that materially affect implementation.
## Suggested prompt structure
Use this structure as a starting point for complex prompts. Keep each section short. Add detail only where it changes behavior.
Role: [the model's function and context]
Personality: [tone and collaboration style]
Goal: [user-visible outcome]
Success criteria: [what must be true before the final answer]
Constraints: [policy, safety, business, evidence, and side-effect limits]
Tools: [which tools to use, when, and what not to use]
Output: [sections, length, format, and tone]
Stop rules: [when to retry, fallback, abstain, ask, or stop]
## Prompt migration workflow
When moving an existing application to GPT-5.6:
1. Switch the model and preserve the current reasoning effort.
2. Run representative evals before changing the prompt.
3. Remove obsolete scaffolding, repeated instructions, and irrelevant tools.
4. Add only the smallest targeted instruction that fixes a measured regression.
5. Re-run evals after each prompt or reasoning change.
Do not rewrite a working prompt stack all at once. Otherwise you cannot tell whether a behavior change came from the model, reasoning setting, prompt, tool set, or runtime.
When a prompt regresses, debug it with a small set of real traces. Identify the failure mode, find the instruction or contradiction that likely caused it, make a surgical edit, and rerun the same cases.

22
.codex-home/skills/.system/openai-docs/references/upgrade-guide.md Normal file
View file

@ -0,0 +1,22 @@
# Model upgrade guidance
Use this file only as a bundled routing fallback when the live migration guide cannot be fetched.
For latest, current, default, or unspecified-model upgrades:
1. Run `scripts/resolve-latest-model-info`.
2. Fetch the returned `migrationGuideUrl` and `promptingGuideUrl` exactly.
3. Treat the live guides as canonical.
4. If remote retrieval fails, disclose that bundled fallback guidance is being used.
For an explicit GPT-5.6 Sol or GPT-5.6-family migration:
1. Preserve the user's explicit target; do not run the latest-model resolver.
2. Fetch the live GPT-5.6 model guidance:
https://developers.openai.com/api/docs/guides/model-guidance?model=gpt-5.6
3. Read `references/upgrading-to-gpt-5p6-sol.md` for skill-specific migration judgment.
4. Read `references/prompting-guide.md` only when prompt changes are needed.
For another explicit model target, preserve that target and fetch its current official guidance. Do not reuse GPT-5.6-specific defaults, API shapes, or compatibility rules for a different model.

448
.codex-home/skills/.system/openai-docs/references/upgrading-to-gpt-5p6-sol.md Normal file
View file

@ -0,0 +1,448 @@
# Upgrading to GPT-5.6 Sol
Use this guide when the user asks to migrate an existing OpenAI API integration, repository, prompt stack, agent, model router, or model picker to GPT-5.6 Sol or the GPT-5.6 family.
The default explicit target is `gpt-5.6-sol`. The alias `gpt-5.6` routes to Sol; use it only when the repository intentionally prefers family aliases. Do not treat every old model usage as a Sol candidate: GPT-5.6 is a family with different cost, latency, context, and quality roles.
Before changing code, retrieve the current live GPT-5.6 model guidance using already-callable official documentation search and fetch, or immediately use official-domain web search and fetch:
https://developers.openai.com/api/docs/guides/model-guidance?model=gpt-5.6
For prompt changes, also read only the `## Prompting Best Practices` section from:
https://developers.openai.com/api/docs/guides/model-guidance?model=gpt-5.6#prompting-best-practices
Treat live docs as canonical for current model IDs, parameters, limits, pricing, and feature availability. This file supplies migration judgment: where to look, what can break, what to preserve, what not to adopt automatically, and how to validate the result.
## Core principle
Do not perform a blind model-string replacement.
First preserve the behavior, latency class, cost class, reasoning level, endpoint contract, tool semantics, cache behavior, and output contract of each usage site. Then make the smallest safe migration. Adopt new GPT-5.6 capabilities only when they solve a measured problem or the user explicitly asks for them.
A model upgrade alone does not authorize adding reasoning fields, changing request schemas, or rewriting tests. Only add explicit reasoning when the old effective behavior is established and omission would change behavior on GPT-5.6.
The main 5.6 migration hazards are:
- choosing Sol for workloads that were intentionally mini, nano, low-cost, or latency-sensitive;
- inheriting 5.6's default `medium` reasoning where the old effective effort was `none`;
- using Chat Completions with function tools without explicitly setting effective reasoning to `none`;
- losing prompt-cache hits when a stable prefix is followed by a changing suffix;
- increasing image or PDF input tokens because omitted or `auto` detail behaves differently;
- applying new cache, persisted-reasoning, Pro, Programmatic Tool Calling, or multi-agent fields to routes that do not support them;
- updating model strings but forgetting registries, allowlists, pricing metadata, capability flags, tests, and UI model pickers.
## Migration posture
Classify every usage site before editing:
1. `simple Sol migration`
- One flagship model usage.
- Same endpoint and request shape can remain.
- Reasoning effort is explicit or its old effective value is known.
- No cache, vision, file, tool, or parser behavior needs implementation changes.
2. `tier-aware family migration`
- The repository exposes multiple model roles, model choices, fallbacks, routers, pricing data, or capability metadata.
- Map each role to Sol, Terra, or Luna instead of replacing everything with Sol.
3. `compatibility migration`
- The safe move requires parameter, endpoint, cache, state, tool-loop, or multimodal-detail changes.
- Make these changes only when implementation work is inside the user's requested scope. Otherwise report the exact blocker and smallest follow-up.
4. `prompt migration`
- The API shape can remain, but representative traces show a prompt-specific regression.
- Make a surgical prompt edit tied to that failure; do not rewrite a working prompt stack wholesale.
- When the task is to update prompting guidance, edit the directly tied prompt surface only. Do not modify runtime request code, model schemas, or tests unless the prompt change requires it.
5. `optional feature adoption`
- Pro mode, persisted reasoning, explicit caching, Programmatic Tool Calling, or multi-agent behavior is being added deliberately.
- Keep this separate from the baseline migration so its effect can be measured.
6. `leave unchanged`
- Historical examples, documentation about old models, snapshots, fixtures, eval baselines, comparison code, intentionally pinned fallbacks, unsupported providers, or ambiguous usages.
When intent is unclear, prefer leaving a usage unchanged and list it for confirmation over silently changing its role.
## Inventory before editing
Search for more than literal model IDs. Inventory:
- model strings, aliases, environment variables, CLI flags, config defaults, and deployment settings;
- SDK calls to Responses, Chat Completions, Batch, or provider adapters;
- reasoning settings, token budgets, sampling settings, and latency timeouts;
- function tools, hosted tools, structured outputs, response parsers, and replay logic;
- system, developer, user, and tool-description prompts tied to each usage;
- routers, fallbacks, model allowlists, enums, regexes, validation schemas, and capability maps;
- model picker UI, display labels, descriptions, context limits, pricing metadata, and provider catalogs;
- prompt-cache keys, retention options, stable-prefix construction, and cache metrics;
- image, PDF, file, OCR, and computer-use inputs;
- tests, fixtures, snapshots, evals, analytics labels, billing tables, and docs.
When changing a default model, search every active default surface: runtime config, environment/config files, setup docs, tests, CLI defaults, and deployment examples. Update them together.
For each usage site, record:
- source model and why it appears to be used;
- endpoint and SDK/client surface;
- prompt surface;
- effective reasoning effort, including defaults;
- latency, cost, context, and quality role;
- tools, structured outputs, caching, state replay, and multimodal inputs;
- downstream parsers or user-visible contracts;
- migration class and validation plan.
## Choose the target model by role
Use this as a starting map, then validate against the repository's workload:
| Existing role | Starting GPT-5.6 target | Reason |
| --- | --- | --- |
| Unsuffixed GPT-5 flagship, GPT-5.5, or GPT-5.4 flagship | `gpt-5.6-sol` | Sol is the flagship-equivalent tier. |
| Mini model, balanced lower-cost route, or medium-throughput worker | `gpt-5.6-terra` | Terra is the mini-like tier. |
| Nano model, classification, extraction, routing, high-volume, or strict-latency route | `gpt-5.6-luna` | Luna is the nano-like tier. |
| GPT-4.1 or GPT-4o latency-sensitive flow | Evaluate Luna and Terra first; use Sol only if quality requires it | A flagship replacement can change latency and cost materially. |
| Reasoning-heavy or hardest quality-first flow | Start with Sol at the old effective effort | Preserve the reasoning contract before tuning. |
| Old Pro usage | Sol plus `reasoning.mode: "pro"`, only if the user wants Pro behavior | GPT-5.6 Pro is a mode, not a separate model slug. |
| Router, fallback, or model picker | Add the family by role | Do not collapse a multi-model design into Sol. |
| Third-party or provider-specific model | Leave unchanged unless the user explicitly requests provider migration | Model-name similarity is not a safe mapping. |
Important limits to check in live docs:
- Sol and Terra have roughly 1.05M context and 128K maximum output.
- Luna has a smaller 400K context and 128K maximum output.
- Sol and Terra long-context requests above 272K input tokens can change pricing for the full request.
Do not invent prices, limits, or capability flags. Fetch them from current docs before updating a registry or UI.
For model pickers and registries, preserve existing model entries by default. Add GPT-5.6 Sol, Terra, and Luna as new options unless the user explicitly asks to replace or remove older models. Do not invent pricing, context limits, capabilities, or metadata unless confirmed from canonical docs.
If using the `gpt-5.6` alias, record the returned `response.model` during validation. Do not assume an alias and an explicit Sol slug appear identically in dashboards, rate-limit configuration, analytics, or billing metadata.
## Preserve effective reasoning before tuning
GPT-5.6 supports `none`, `low`, `medium`, `high`, `xhigh`, and `max`. If omitted, GPT-5.6 defaults to `medium`.
This is a behavioral migration hazard:
- GPT-5.5 commonly defaulted to `medium`.
- GPT-5.4, mini, and nano usages commonly defaulted to `none`.
- A previously omitted setting can therefore become slower, more expensive, and incompatible with Chat Completions function tools after the model swap.
For each usage:
1. If effort is explicit, preserve it for the first 5.6 run when supported.
2. If effort is omitted and the old effective default is known, add it explicitly only when GPT-5.6's omitted default would change behavior. If both old and new omitted defaults are the same, keep it omitted.
3. If the old effective value is unknown, do not guess. Flag it and compare the old behavior with 5.6 at the likely baseline.
4. After the baseline passes, test the same setting and one lower on representative tasks.
5. Use `xhigh` or `max` only for hard quality-first workloads where evals show a meaningful gain.
Do not globally recommend `max`. Before increasing effort, check whether the actual failure is a missing success criterion, dependency rule, tool-routing rule, state-replay bug, or validation loop.
Use the field shape that belongs to the endpoint.
Responses:
```json
{
"model": "gpt-5.6-sol",
"reasoning": { "effort": "none" }
}
```
Chat Completions:
```json
{
"model": "gpt-5.6-sol",
"reasoning_effort": "none"
}
```
## Chat Completions and function tools
This is the most important endpoint-specific check.
For GPT-5.6, function tools in Chat Completions are compatible only with effective reasoning `none`. Reasoning with tools should use the Responses API.
Because GPT-5.6 defaults to `medium`, this combination is unsafe:
```json
{
"model": "gpt-5.6-luna",
"tools": [{ "type": "function", "function": { "...": "..." } }]
}
```
For a latency-sensitive Chat Completions flow that must keep function tools, explicitly preserve `none`:
```json
{
"model": "gpt-5.6-luna",
"reasoning_effort": "none",
"tools": [{ "type": "function", "function": { "...": "..." } }]
}
```
If the application needs both reasoning and tools:
- migrate that flow to Responses when implementation changes are in scope;
- otherwise report it as a compatibility blocker;
- do not hide the incompatibility by removing tools, dropping required reasoning, or changing the workload's behavior without approval.
If the live API rejects the intended `none` path, treat it as a current API compatibility issue and report the exact request and error rather than inventing a workaround.
## Responses API and conversation state
Prefer Responses for reasoning, tools, multi-turn agents, and new 5.6 capabilities.
For ordinary multi-turn Responses calls, preserve the repository's existing state strategy. Do not add persisted reasoning merely because it exists.
If deliberately enabling persisted reasoning:
- use `reasoning.context: "all_turns"` only when the objective and assumptions remain stable;
- prefer `previous_response_id` when the server can carry state;
- when replaying manually, preserve every prior user input and every relevant output item, not only assistant text;
- with `store: false` or ZDR, request and replay `reasoning.encrypted_content`;
- use current-turn behavior when old reasoning may be stale or misleading.
For manual replay, preserve item types, IDs, call IDs, caller metadata, and assistant phase values exactly. Incomplete replay can silently reduce quality or break tool continuation.
## Prompt caching
Do not assume old cache-hit behavior survives the model swap.
GPT-5.6 implicit caching places a managed breakpoint near the latest user or tool message and no longer relies on 128-token rounding. A prompt with a large stable prefix followed by a changing suffix can therefore lose cache hits even when the stable prefix itself has not changed.
Audit:
- large reusable system/developer prompts;
- dynamic suffixes appended to otherwise stable prompts;
- changing timestamps, request IDs, user-specific values, or tool lists in the prefix;
- cache keys, retention settings, and cache dashboards;
- token accounting that assumes reads only and ignores writes.
Migration rules:
- keep reusable prefixes stable;
- do not churn large system prompts unnecessarily;
- compare old and new `cached_tokens`, `cache_write_tokens`, latency, and cost;
- use explicit cache breakpoints only when a measured workload has a stable boundary that implicit caching misses;
- do not globally convert every prompt to explicit caching;
- do not send 5.6-only cache fields to older routes in a mixed-model system.
When old and GPT-5.6 routes share a request builder, isolate GPT-5.6-only fields instead of applying them globally.
The new top-level request shape uses `prompt_cache_options`, for example:
```json
{
"prompt_cache_options": {
"mode": "explicit",
"ttl": "30m"
}
}
```
Place explicit breakpoints at the actual stable rendered boundary using `prompt_cache_breakpoint`. Preserve `prompt_cache_key` when the application already uses it. Treat the older `prompt_cache_retention` shape as deprecated and verify the live docs before rewriting it.
Cache writes cost more than ordinary uncached input, so a lower hit rate can be both slower and more expensive.
## Images, PDFs, files, and long context
GPT-5.6 can change token and latency behavior without any prompt change:
- for image inputs, omitted or `auto` image detail can preserve original dimensions;
- for PDF/file inputs in Responses, omitted or `input_file.detail: "auto"` can use high page-image detail;
- Chat Completions file inputs do not expose the same detail control;
- long-context Sol and Terra requests can cross pricing thresholds;
- Luna's smaller context can break workloads that fit in Sol or Terra.
For multimodal or long-context usages:
1. Measure input tokens and latency before and after.
2. Make detail explicit when cost or latency matters.
3. Resize images or use lower detail when the task does not need original spatial precision.
4. Keep original/high detail for dense, coordinate-sensitive, OCR, localization, or visual-inspection tasks where it materially improves quality.
5. Test worst-case context lengths, not only typical requests.
Do not claim a capability was removed based only on a missing metadata flag. Verify against current docs and a representative request.
## Structured outputs, parsers, and tool contracts
Keep output contracts explicit:
- preserve JSON schemas, required fields, enums, refusal handling, and parser expectations;
- preserve tool names, parameter schemas, call IDs, and retry behavior;
- keep citations, evidence fields, or native artifacts when downstream consumers require them;
- validate that the final answer still satisfies the contract, not merely that a tool call succeeded.
Do not fix a failing migration by weakening a schema, deleting required behavior, removing routes, dropping tools, or changing business logic unless the user explicitly asked for that product change.
## Optional: Pro mode
Do not enable Pro mode during a baseline migration unless the old usage was Pro-like or the user explicitly asks for it.
GPT-5.6 Pro uses the base model with a reasoning mode:
```json
{
"model": "gpt-5.6-sol",
"reasoning": {
"mode": "pro",
"effort": "medium"
}
}
```
Rules:
- use Responses, not Chat Completions;
- do not search for or invent a separate `gpt-5.6-pro` slug;
- supported Pro efforts begin at `medium`;
- mode and effort are separate decisions;
- compare task quality, total latency, and actual billed token usage against standard mode.
If migrating a legacy Pro slug, make the mode change explicit and evaluate it separately from ordinary Sol migration.
## Optional: Programmatic Tool Calling
Programmatic Tool Calling is not a required part of moving to GPT-5.6. Add it only when code can reduce large structured intermediate results before they return to model context.
Good candidates:
- bounded read-only filtering, joining, sorting, ranking, deduplication, and aggregation;
- batching many similar records;
- repeated deterministic validation;
- map-reduce style retrieval with a compact result schema.
Poor candidates:
- one direct tool call;
- adaptive workflows where each result changes the next decision;
- write, approval, or side-effecting flows;
- citation-heavy or native-artifact flows;
- semantic judgment that should remain visible to the model.
Request-shape requirements:
```json
{
"tools": [
{ "type": "programmatic_tool_calling" },
{
"type": "function",
"name": "lookup_records",
"allowed_callers": ["programmatic"]
}
]
}
```
Do not nest `programmatic_tool_calling` under another `tools` property. When enabled, the host must handle `program`, program-issued `function_call`, `function_call_output`, and `program_output` items. Preserve the original `call_id` and `caller` when returning function results.
Constrain the stage, eligible read-only tools, output schema, retry limit, and handoff back to direct judgment. Validate the final user-visible answer; a correct program result can still become an incorrect final answer.
## Optional: multi-agent beta
Do not enable multi-agent behavior during a baseline migration unless the application already has a clear parallelizable workflow and the user asks for it.
Enabling it requires:
- the `OpenAI-Beta: responses_multi_agent=v1` header;
- `multi_agent: { "enabled": true, "max_concurrent_subagents": 3 }`;
- handling `multi_agent_call`, `multi_agent_call_output`, and `agent_message` items;
- executing ordinary developer-defined function calls from any agent and returning all required outputs;
- preserving new items for replay and tracing;
- checking incompatibilities with compaction, reasoning summaries, and tool-call limits in current docs.
Cap concurrency. Do not let a migration task create unbounded subagents, duplicate work, or finish without a final synthesis.
## Prompt migration judgment
After the model and API baseline is working, run representative traces before editing prompts. Change prompts only for measured failures.
For GPT-5.6, prefer:
- shorter, outcome-oriented prompts;
- explicit success criteria, dependencies, stopping conditions, and completion boundaries;
- preserved user-provided values;
- decision criteria for implicit choices instead of universal defaults or keyword maps;
- explicit autonomy and permission boundaries;
- explicit tool routing, resource links, breadcrumbs, and expected tool choice;
- staged plans, current-layer awareness, and concise handoffs for long work;
- real validation before declaring completion.
Avoid:
- generic `be brief`, `be thorough`, or `think step by step` instructions;
- blanket language instructions that can cause unwanted language switching;
- repeating `ask first` until safe local work becomes blocked;
- giant prompt rewrites that make the source of a regression impossible to identify;
- telling the model to minimize tool loops when correctness, evidence, or required validation needs more work.
For coding or agentic migrations, add concrete preservation and verification rules:
```
Preserve existing functionality, routes, outputs, and user-visible behavior.
Do not delete or disable required behavior merely to make the build pass.
Before finishing, run the relevant build, tests, type checks, render or smoke
checks, and report the evidence.
```
For long-running work, define the current layer: research, design, implementation, review, or external coordination. Do not let the model silently move to another layer.
## Upgrade workflow
1. Fetch current live 5.6 docs and the Prompting Best Practices section.
2. Inventory every usage site and its adjacent prompt, config, registry, parser, and test surfaces.
3. Classify each usage by role and migration class.
4. Choose Sol, Terra, or Luna by the existing workload's role.
5. Preserve the old effective reasoning effort explicitly.
6. Run the compatibility gates:
- endpoint and SDK support;
- Chat Completions plus function tools;
- cache topology and cache fields;
- context length and long-context cost;
- image, PDF, and file detail;
- structured outputs and parsers;
- Responses state replay and tool continuation;
- mixed-model routing and unsupported new fields.
7. Apply the smallest safe model, config, registry, and prompt changes.
8. Do not add optional Pro, persisted reasoning, PTC, explicit caching, or multi-agent behavior unless needed and measurable.
9. Run existing tests and representative evals.
10. Report changed, unchanged, blocked, and confirmation-needed sites separately.
## Validation matrix
Prefer a controlled comparison:
1. old model + old prompt + old settings;
2. GPT-5.6 target + same prompt + preserved effective reasoning;
3. GPT-5.6 target + same prompt + one lower effort;
4. GPT-5.6 target + the smallest prompt or API fix required by a measured failure;
5. optional feature treatment, isolated from the baseline.
Measure what matters for the workflow:
- task success and user-visible quality;
- structured-output validity and parser success;
- tool choice, tool arguments, retries, loop count, and completion rate;
- TTFT, end-to-end latency, timeout rate, and concurrency behavior;
- input, output, reasoning, cached, and cache-write tokens;
- cost per successful task;
- long-context, compaction, and replay behavior;
- image/PDF token use and visual/OCR accuracy;
- completeness, preserved behavior, citations, and validation evidence.
For model routers and pickers, test at least one representative workload for each role. Verify that the cheapest or fastest tier is not accidentally used for quality-critical work and that Sol is not accidentally used for every workload.
## Required final report
Return:
- `Current usage inventory`: each model site, endpoint, role, prompt surface, and old effective reasoning.
- `Target mapping`: Sol, Terra, Luna, unchanged, or confirmation-needed, with the reason.
- `Changes made`: model strings, reasoning settings, prompts, registries, metadata, tests, and API-shape changes.
- `Compatibility checks`: Chat Completions/tools, caching, state replay, multimodal detail, context/cost, schemas, and mixed-model routing.
- `Prompt changes`: each surgical edit and the failure mode it addresses.
- `Validation`: commands, evals, traces, before/after measurements, and remaining gaps.
- `Unchanged sites`: historical, pinned, ambiguous, or intentionally role-specific usages.
- `Blockers and open questions`: exact issue, why it is unsafe to guess, and the smallest next step.
Never say the migration is complete merely because model strings changed. It is complete only when the affected behavior and contracts have been validated or the remaining gaps are stated explicitly.

598
.codex-home/skills/.system/openai-docs/scripts/fetch-codex-manual.mjs Normal file
View file

@ -0,0 +1,598 @@
#!/usr/bin/env node
import {
access,
mkdir,
readFile,
rename,
rm,
stat,
writeFile,
} from "node:fs/promises";
import { constants as fsConstants } from "node:fs";
import { execFile } from "node:child_process";
import { createHash } from "node:crypto";
import path from "node:path";
import process from "node:process";
import { pathToFileURL } from "node:url";
import { inspect, promisify } from "node:util";
const DEFAULT_MANUAL_URL = "https://developers.openai.com/codex/codex-manual.md";
const DEFAULT_CACHE_DIR_NAME = "openai-docs-cache";
const CACHE_FILE_NAME = "codex-manual.md";
const OUTLINE_FILE_NAME = "codex-manual.outline.md";
const HASH_HEADER = "x-content-sha256";
const USER_AGENT = "codex-openai-docs";
const execFileAsync = promisify(execFile);
class ManualFetchError extends Error {
constructor(message, options) {
super(message, options);
this.name = "ManualFetchError";
}
}
const sha256 = (value) => createHash("sha256").update(value).digest("hex");
const withTimeout = async (promiseFactory, timeoutMs) => {
const controller = new AbortController();
const timeout = setTimeout(() => controller.abort(), timeoutMs);
try {
return await promiseFactory(controller.signal);
} finally {
clearTimeout(timeout);
}
};
const proxyConfigured = () =>
process.env.HTTP_PROXY ||
process.env.HTTPS_PROXY ||
process.env.http_proxy ||
process.env.https_proxy;
const responseHeaders = (headers) => ({
get(name) {
return headers.get(name.toLowerCase()) ?? null;
},
});
const makeResponse = ({ body, headers, status }) => ({
headers: responseHeaders(headers),
ok: status >= 200 && status < 300,
status,
async text() {
return body;
},
});
const parseCurlHeaders = (rawHeaders) => {
const normalized = rawHeaders.replace(/\r\n/g, "\n").trim();
const blocks = normalized.split(/\n\n+/).filter(Boolean);
const headerBlock = [...blocks]
.reverse()
.find((block) => block.startsWith("HTTP/"));
if (!headerBlock) {
throw new ManualFetchError("curl did not return HTTP response headers.");
}
const [statusLine, ...lines] = headerBlock.split("\n");
const statusMatch = /^HTTP\/\S+\s+(\d{3})/.exec(statusLine);
if (!statusMatch) {
throw new ManualFetchError(
`Could not parse HTTP status from curl response: ${statusLine}`
);
}
const headers = new Map();
lines.forEach((line) => {
const separator = line.indexOf(":");
if (separator === -1) return;
const name = line.slice(0, separator).trim().toLowerCase();
const value = line.slice(separator + 1).trim();
headers.set(name, value);
});
return {
headers,
status: Number(statusMatch[1]),
};
};
const tempFilePath = (cacheDir, suffix) =>
path.join(
cacheDir,
`.fetch-codex-manual-${process.pid}-${Date.now()}-${Math.random()
.toString(16)
.slice(2)}${suffix}`
);
const requestManualWithCurl = async (url, { cacheDir, method, timeoutMs }) => {
const headerPath = tempFilePath(cacheDir, ".headers");
const bodyPath = tempFilePath(cacheDir, ".body");
const curlNames =
process.platform === "win32" ? ["curl.exe", "curl"] : ["curl"];
const args = [
"--silent",
"--show-error",
"--location",
"--dump-header",
headerPath,
"--output",
bodyPath,
"--user-agent",
USER_AGENT,
"--max-time",
String(Math.max(1, Math.ceil(timeoutMs / 1000))),
];
if (method === "HEAD") {
args.push("--head");
} else {
args.push("--request", method);
}
args.push(url);
let lastError;
for (const curlName of curlNames) {
try {
await execFileAsync(curlName, args, { windowsHide: true });
const [rawHeaders, body] = await Promise.all([
readFile(headerPath, "utf8"),
readFile(bodyPath, "utf8"),
]);
const { headers, status } = parseCurlHeaders(rawHeaders);
return makeResponse({ body, headers, status });
} catch (error) {
lastError = error;
if (error?.code !== "ENOENT") break;
} finally {
await Promise.all([
rm(headerPath, { force: true }),
rm(bodyPath, { force: true }),
]);
}
}
if (lastError?.code === "ENOENT") {
throw new ManualFetchError("curl is unavailable in this environment.", {
cause: lastError,
});
}
throw new ManualFetchError(`${method} ${url} could not be fetched.`, {
cause: lastError,
});
};
const requestManualWithFetch = async (url, { method, timeoutMs }) => {
if (typeof fetch !== "function") {
throw new ManualFetchError(
"Native fetch is unavailable in this Node runtime."
);
}
return withTimeout(
(signal) =>
fetch(url, {
method,
headers: { "User-Agent": USER_AGENT },
signal,
}),
timeoutMs
);
};
const requestManual = async (url, { cacheDir, method, timeoutMs }) => {
const preferCurl = Boolean(proxyConfigured()) || typeof fetch !== "function";
const transports = preferCurl
? [
() => requestManualWithCurl(url, { cacheDir, method, timeoutMs }),
() => requestManualWithFetch(url, { method, timeoutMs }),
]
: [
() => requestManualWithFetch(url, { method, timeoutMs }),
() => requestManualWithCurl(url, { cacheDir, method, timeoutMs }),
];
let lastError;
for (const transport of transports) {
try {
const response = await transport();
if (!response.ok) {
throw new ManualFetchError(
`${method} ${url} failed with HTTP ${response.status}.`
);
}
return response;
} catch (error) {
lastError = error;
}
}
throw new ManualFetchError(`${method} ${url} could not be fetched.`, {
cause: lastError,
});
};
const readHeaderSha = (response) => {
const value = response.headers.get(HASH_HEADER);
if (!value || !/^[a-f0-9]{64}$/i.test(value)) {
throw new ManualFetchError(`Manual response is missing ${HASH_HEADER}.`);
}
return value.toLowerCase();
};
const nearestExistingParent = async (target) => {
let current = target;
while (true) {
try {
const info = await stat(current);
return info.isDirectory() ? current : null;
} catch (error) {
if (error?.code !== "ENOENT") return null;
}
const parent = path.dirname(current);
if (parent === current) return null;
current = parent;
}
};
const usableCacheDir = async (cacheDir) => {
if (!cacheDir) return null;
const resolved = path.resolve(cacheDir);
try {
const info = await stat(resolved);
if (!info.isDirectory()) return null;
} catch (error) {
if (error?.code !== "ENOENT") return null;
}
const parent = await nearestExistingParent(resolved);
if (!parent) return null;
try {
await access(parent, fsConstants.W_OK | fsConstants.X_OK);
} catch {
return null;
}
return resolved;
};
const defaultCacheDirCandidates = () => {
const candidates = [];
const seen = new Set();
const pushCandidate = (candidate) => {
if (!candidate || seen.has(candidate)) return;
seen.add(candidate);
candidates.push(candidate);
};
[process.env.TMPDIR, process.env.TEMP, process.env.TMP].forEach((baseDir) => {
if (baseDir) {
pushCandidate(path.join(baseDir, DEFAULT_CACHE_DIR_NAME));
}
});
if (process.platform !== "win32") {
pushCandidate(`/private/tmp/${DEFAULT_CACHE_DIR_NAME}`);
pushCandidate(`/tmp/${DEFAULT_CACHE_DIR_NAME}`);
}
return candidates;
};
const resolveCacheDir = async (cacheDir) => {
if (cacheDir) {
return usableCacheDir(cacheDir);
}
for (const candidate of defaultCacheDirCandidates()) {
const usable = await usableCacheDir(candidate);
if (usable) return usable;
}
return null;
};
const cacheFilePath = (cacheDir) => path.join(cacheDir, CACHE_FILE_NAME);
const outlineFilePath = (cacheDir) => path.join(cacheDir, OUTLINE_FILE_NAME);
const manualLines = (manual) => {
const lines = manual.replace(/\r\n/g, "\n").split("\n");
if (lines[lines.length - 1] === "") lines.pop();
return lines;
};
const sectionTitle = (rawTitle) =>
rawTitle.replace(/\s+#+\s*$/, "").replace(/\s+/g, " ").trim();
const buildOutline = (manual) => {
const lines = manualLines(manual);
const headings = [];
let inFence = false;
lines.forEach((line, index) => {
if (/^\s*(```|~~~)/.test(line)) {
inFence = !inFence;
return;
}
if (inFence) return;
const match = /^(#{1,6})\s+(.+?)\s*$/.exec(line);
if (!match) return;
const level = match[1].length;
if (level < 2 || level > 3) return;
headings.push({
level,
title: sectionTitle(match[2]),
startLine: index + 1,
endLine: lines.length,
});
});
for (let index = 0; index < headings.length; index += 1) {
const heading = headings[index];
const nextPeer = headings
.slice(index + 1)
.find((candidate) => candidate.level <= heading.level);
if (nextPeer) {
heading.endLine = nextPeer.startLine - 1;
}
}
if (headings.length === 0) {
return {
headingCount: 0,
lineCount: lines.length,
text: "No markdown headings found.",
};
}
const minLevel = Math.min(...headings.map((heading) => heading.level));
return {
headingCount: headings.length,
lineCount: lines.length,
text: headings
.map((heading) => {
const indent = " ".repeat(heading.level - minLevel);
return `${indent}- ${heading.title} (lines ${heading.startLine}-${heading.endLine})`;
})
.join("\n"),
};
};
const outlineMarkdown = (outline) => `# Codex Manual Outline\n\n${outline.text}\n`;
const manualStatusLine = (status) =>
status.cacheStatus === "hit"
? "Manual status: local manual was already current."
: "Manual status: local manual was updated.";
const formatResult = ({ status, outlineText }) =>
[
`Manual path: ${status.manualPath}`,
`Outline path: ${status.outlinePath}`,
manualStatusLine(status),
"",
outlineText,
].join("\n");
const readCachedManual = async (cacheDir, expectedSha256) => {
try {
const manual = await readFile(cacheFilePath(cacheDir), "utf8");
return sha256(manual) === expectedSha256 ? manual : null;
} catch {
return null;
}
};
const writeCachedManual = async (cacheDir, manual) => {
await mkdir(cacheDir, { recursive: true });
const tmpPath = tempFilePath(cacheDir, `.${CACHE_FILE_NAME}.tmp`);
await writeFile(tmpPath, manual, "utf8");
await rename(tmpPath, cacheFilePath(cacheDir));
};
const writeOutline = async (cacheDir, outlineText) => {
await mkdir(cacheDir, { recursive: true });
const tmpPath = tempFilePath(cacheDir, `.${OUTLINE_FILE_NAME}.tmp`);
await writeFile(tmpPath, outlineText, "utf8");
await rename(tmpPath, outlineFilePath(cacheDir));
};
const fetchCodexManual = async ({
manualUrl = DEFAULT_MANUAL_URL,
cacheDir,
timeoutMs = 30000,
} = {}) => {
const resolvedCacheDir = await resolveCacheDir(cacheDir);
if (!resolvedCacheDir) {
throw new ManualFetchError(
"Manual cache directory is unavailable; pass --cache-dir to override or use OpenAI Docs MCP fallback."
);
}
await mkdir(resolvedCacheDir, { recursive: true });
const headResponse = await requestManual(manualUrl, {
cacheDir: resolvedCacheDir,
method: "HEAD",
timeoutMs,
});
const expectedSha256 = readHeaderSha(headResponse);
const manualPath = cacheFilePath(resolvedCacheDir);
const outlinePath = outlineFilePath(resolvedCacheDir);
const checkedAt = new Date().toISOString();
const cachedManual = await readCachedManual(resolvedCacheDir, expectedSha256);
if (cachedManual !== null) {
const outline = buildOutline(cachedManual);
const outlineText = outlineMarkdown(outline);
await writeOutline(resolvedCacheDir, outlineText);
return {
outlineText,
status: {
manualUrl,
headerSha256: expectedSha256,
fetchedManualSha256: expectedSha256,
manualHashMatches: true,
cacheStatus: "hit",
cacheDir: resolvedCacheDir,
manualPath,
outlinePath,
checkedAt,
lineCount: outline.lineCount,
headingCount: outline.headingCount,
},
};
}
const getResponse = await requestManual(manualUrl, {
cacheDir: resolvedCacheDir,
method: "GET",
timeoutMs,
});
const getHeaderSha256 = readHeaderSha(getResponse);
if (getHeaderSha256 !== expectedSha256) {
throw new ManualFetchError(
`${HASH_HEADER} changed between HEAD and GET for ${manualUrl}.`
);
}
const manualText = await getResponse.text();
const actualSha256 = sha256(manualText);
const manualHashMatches = actualSha256 === expectedSha256;
if (!manualHashMatches) {
throw new ManualFetchError(
`${HASH_HEADER} did not match the fetched manual body for ${manualUrl}.`
);
}
await writeCachedManual(resolvedCacheDir, manualText);
const outline = buildOutline(manualText);
const outlineText = outlineMarkdown(outline);
await writeOutline(resolvedCacheDir, outlineText);
return {
outlineText,
status: {
manualUrl,
headerSha256: expectedSha256,
fetchedManualSha256: actualSha256,
manualHashMatches,
cacheStatus: "updated",
cacheDir: resolvedCacheDir,
manualPath,
outlinePath,
checkedAt,
lineCount: outline.lineCount,
headingCount: outline.headingCount,
},
};
};
const parseArgs = (argv) => {
const args = {
manualUrl: DEFAULT_MANUAL_URL,
cacheDir: undefined,
timeoutMs: 30000,
statusJson: false,
};
for (let index = 0; index < argv.length; index += 1) {
const arg = argv[index];
if (arg === "--manual-url") {
args.manualUrl = argv[++index];
} else if (arg === "--cache-dir") {
args.cacheDir = argv[++index];
} else if (arg === "--timeout-ms") {
args.timeoutMs = Number(argv[++index]);
} else if (arg === "--status-json") {
args.statusJson = true;
} else {
throw new ManualFetchError(`Unknown argument: ${arg}`);
}
}
if (!args.manualUrl) {
throw new ManualFetchError("--manual-url cannot be empty.");
}
if (!Number.isFinite(args.timeoutMs) || args.timeoutMs <= 0) {
throw new ManualFetchError("--timeout-ms must be a positive number.");
}
return args;
};
const main = async () => {
const args = parseArgs(process.argv.slice(2));
const { outlineText, status } = await fetchCodexManual(args);
process.stdout.write(formatResult({ status, outlineText }));
if (args.statusJson) {
console.error(JSON.stringify(status));
}
};
const envProxyHint = () => {
if (proxyConfigured()) {
return "Hint: proxy env vars are present. This helper prefers `curl` in proxied sessions; if requests still fail, verify `curl` is installed and the proxy configuration is valid.";
}
if (typeof fetch !== "function") {
return "Hint: native fetch is unavailable in this Node runtime. Install `curl` or use a newer Node version to fetch the manual.";
}
if (process.platform === "win32") {
return "Hint: on Windows, pass a cache dir under `%TEMP%` or `%TMP%`.";
}
return null;
};
const formatErrorDetails = (error) => {
const details = inspect(error, {
breakLength: 120,
colors: false,
compact: false,
depth: 8,
});
if (!error?.cause) {
return details;
}
return `${details}\n\nCause:\n${inspect(error.cause, {
breakLength: 120,
colors: false,
compact: false,
depth: 8,
})}`;
};
const isCliEntrypoint = () => {
const entrypoint = process.argv[1];
if (!entrypoint) {
return false;
}
return pathToFileURL(entrypoint).href === import.meta.url;
};
if (isCliEntrypoint()) {
main().catch((error) => {
console.error(`Error: ${error.message}`);
const hint = envProxyHint();
if (hint) {
console.error(hint);
}
console.error("");
console.error("Details:");
console.error(formatErrorDetails(error));
process.exitCode = 1;
});
}
export { DEFAULT_MANUAL_URL, fetchCodexManual };

39
.codex-home/skills/.system/openai-docs/scripts/resolve-latest-model-info Normal file
View file

@ -0,0 +1,39 @@
#!/bin/sh
set -eu
SCRIPT_DIR=$(CDPATH= cd -- "$(dirname -- "$0")" && pwd)
SCRIPT_PATH="$SCRIPT_DIR/resolve-latest-model-info.cjs"
is_compatible_node() {
"$1" -e 'const major = Number(process.versions.node.split(".")[0]); process.exit(major >= 18 ? 0 : 1)' >/dev/null 2>&1
}
run_if_compatible() {
CANDIDATE=$1
shift
if [ -n "$CANDIDATE" ] && [ -x "$CANDIDATE" ] && is_compatible_node "$CANDIDATE"; then
exec "$CANDIDATE" "$SCRIPT_PATH" "$@"
fi
}
if [ -n "${NODE:-}" ]; then
run_if_compatible "$NODE" "$@"
fi
PATH_NODE=$(command -v node 2>/dev/null || true)
if [ -n "$PATH_NODE" ]; then
run_if_compatible "$PATH_NODE" "$@"
fi
for CANDIDATE in \
"$HOME/.cache/codex-runtimes/codex-primary-runtime/dependencies/node/bin/node" \
"$HOME/.cache/codex-runtimes/codex-primary-runtime/dependencies/bin/node" \
"/opt/homebrew/bin/node" \
"/usr/local/bin/node" \
"/usr/bin/node"
do
run_if_compatible "$CANDIDATE" "$@"
done
echo "No usable Node.js 18+ runtime found for resolve-latest-model-info.cjs" >&2
exit 127

165
.codex-home/skills/.system/openai-docs/scripts/resolve-latest-model-info.cjs Normal file
View file

@ -0,0 +1,165 @@
#!/usr/bin/env node
// Keep this entrypoint CommonJS-safe when the skill is copied into a type=module repo.
const fs = require("node:fs/promises");
const path = require("node:path");
const DEFAULT_URL =
"https://developers.openai.com/api/docs/guides/latest-model.md";
const DEFAULT_BASE_URL = "https://developers.openai.com";
function parseArgs(argv) {
const args = {
source: process.env.LATEST_MODEL_URL || DEFAULT_URL,
baseUrl: process.env.LATEST_MODEL_BASE_URL || DEFAULT_BASE_URL,
};
for (let i = 2; i < argv.length; i += 1) {
const arg = argv[i];
if (arg === "--source" || arg === "--url") {
args.source = argv[i + 1];
i += 1;
} else if (arg === "--base-url") {
args.baseUrl = argv[i + 1];
i += 1;
}
}
return args;
}
async function readSource(source) {
if (source.startsWith("file://")) {
return fs.readFile(new URL(source), "utf8");
}
if (!/^https?:\/\//.test(source)) {
return fs.readFile(path.resolve(source), "utf8");
}
let lastError;
for (let attempt = 1; attempt <= 3; attempt += 1) {
try {
const response = await fetch(source, {
headers: { accept: "text/markdown,text/plain,*/*" },
});
if (response.ok) {
return response.text();
}
lastError = new Error("failed to fetch " + source + ": " + response.status);
if (response.status < 500 && response.status !== 429) {
break;
}
} catch (error) {
lastError = error;
}
if (attempt < 3) {
await new Promise((resolve) => setTimeout(resolve, 250 * attempt));
}
}
throw lastError;
}
function parseIndentedInfo(lines, startIndex) {
const info = {};
for (let i = startIndex + 1; i < lines.length; i += 1) {
const line = lines[i];
if (!line.trim()) {
continue;
}
const match = line.match(/^ {2}([A-Za-z][A-Za-z0-9_-]*):\s*(.+?)\s*$/);
if (!match) {
break;
}
info[match[1]] = match[2].replace(/^["']|["']$/g, "");
}
return info;
}
function parseFlatInfo(block) {
const info = {};
for (const line of block.split(/\r?\n/)) {
const match = line.match(/^\s*([A-Za-z][A-Za-z0-9_-]*):\s*(.+?)\s*$/);
if (match) {
info[match[1]] = match[2].replace(/^["']|["']$/g, "");
}
}
return info;
}
function extractLatestModelInfo(markdown) {
const lines = markdown.split(/\r?\n/);
const latestModelInfoIndex = lines.findIndex((line) =>
/^latestModelInfo:\s*$/.test(line)
);
if (latestModelInfoIndex >= 0) {
return parseIndentedInfo(lines, latestModelInfoIndex);
}
const commentMatch = markdown.match(
/<!--\s*latestModelInfo\s*\n([\s\S]*?)\n\s*-->/m
);
if (commentMatch) {
return parseFlatInfo(commentMatch[1]);
}
return undefined;
}
function modelToSkillSlug(model) {
return model.trim().replace(/\./g, "p");
}
function absoluteUrl(baseUrl, value) {
return new URL(value, baseUrl).toString();
}
function normalizeInfo(info, baseUrl) {
const model = info?.model?.trim();
const migrationGuide = info?.migrationGuide?.trim();
const promptingGuide = info?.promptingGuide?.trim();
if (!model || !migrationGuide || !promptingGuide) {
throw new Error(
"latestModelInfo must include model, migrationGuide, and promptingGuide"
);
}
return {
model,
modelSlug: modelToSkillSlug(model),
migrationGuideUrl: absoluteUrl(baseUrl, migrationGuide),
promptingGuideUrl: absoluteUrl(baseUrl, promptingGuide),
};
}
async function main() {
const { source, baseUrl } = parseArgs(process.argv);
const markdown = await readSource(source);
const info = extractLatestModelInfo(markdown);
if (!info) {
throw new Error(`latestModelInfo block not found in ${source}`);
}
process.stdout.write(
`${JSON.stringify(normalizeInfo(info, baseUrl), null, 2)}\n`
);
}
main().catch((error) => {
console.error(error.message);
process.exit(1);
});

243
.codex-home/skills/.system/plugin-creator/SKILL.md Normal file
View file

@ -0,0 +1,243 @@
---
name: plugin-creator
description: Create and scaffold plugin directories for Codex with a required `.codex-plugin/plugin.json`, optional plugin folders/files, valid manifest defaults, and personal-marketplace entries by default. Use when Codex needs to create a new personal plugin, add optional plugin structure, generate or update marketplace entries for plugin ordering and availability metadata, or update an existing local plugin during development with the CLI-driven cachebuster and reinstall flow.
---
# Plugin Creator
## Quick Start
1. Run the scaffold script:
```bash
# Plugin names are normalized to lower-case hyphen-case and must be <= 64 chars.
# The generated folder and plugin.json name are always the same.
# Run from the skill root (the directory containing this `SKILL.md`).
# By default creates in `~/plugins/<plugin-name>`.
python3 scripts/create_basic_plugin.py <plugin-name>
```
2. Edit `<plugin-path>/.codex-plugin/plugin.json` when the request gives specific metadata.
The scaffold starts with valid defaults and must not contain `[TODO: ...]` placeholders.
3. Generate or update the personal marketplace entry when the plugin should appear in Codex UI ordering:
```bash
# Personal marketplace entries default to `~/.agents/plugins/marketplace.json`.
python3 scripts/create_basic_plugin.py my-plugin --with-marketplace
```
Only specify `--marketplace-name <name>` when the default `personal` marketplace name is already
taken or installed and you need to seed a different new marketplace file:
```bash
python3 scripts/create_basic_plugin.py my-plugin \
--with-marketplace \
--marketplace-name team-local
```
Only use a repo/team marketplace when the user specifically asks for that destination:
```bash
python3 scripts/create_basic_plugin.py my-plugin \
--path <repo-root>/plugins \
--marketplace-path <repo-root>/.agents/plugins/marketplace.json \
--with-marketplace
```
When the user specifies a marketplace path, make sure that marketplace is actually installed before
telling the user to reinstall from it. The default personal marketplace file at
`~/.agents/plugins/marketplace.json` is discovered implicitly, but other marketplace paths are not.
On Windows, use the equivalent path under the user profile.
4. Generate/adjust optional companion folders as needed:
```bash
python3 scripts/create_basic_plugin.py my-plugin \
--path <parent-plugin-directory> \
--marketplace-path <marketplace-json-path> \
--with-skills --with-hooks --with-scripts --with-assets --with-mcp --with-apps --with-marketplace
```
`<parent-plugin-directory>` is the directory where the plugin folder `<plugin-name>` will be
created (for example `~/plugins`).
5. Before handing back a generated plugin, run:
```bash
python3 scripts/validate_plugin.py <plugin-path>
```
For updates to an existing local plugin during development, keep the scaffold flow as-is and use the
reference instead of hand-editing marketplace files:
```bash
python3 scripts/update_plugin_cachebuster.py <plugin-path>
```
Prefer the helper default cachebuster unless the user explicitly asks for a specific override.
See `references/installing-and-updating.md` for the expected cachebuster and reinstall flow while iterating on an existing local plugin.
## What this skill creates
- Default marketplace-backed scaffolds use the personal marketplace file at
`~/.agents/plugins/marketplace.json`, with plugins generally being stored in
`~/plugins/<plugin-name>/`.
- Creates plugin root at `/<parent-plugin-directory>/<plugin-name>/`.
- Always creates `/<parent-plugin-directory>/<plugin-name>/.codex-plugin/plugin.json`.
- Fills the manifest with the validated schema shape that the ingestion path accepts.
- Creates or updates `~/.agents/plugins/marketplace.json` when `--with-marketplace` is set.
- If the marketplace file does not exist yet, seed a personal marketplace root before adding the first plugin entry.
- `<plugin-name>` is normalized using skill-creator naming rules:
- `My Plugin` → `my-plugin`
- `My--Plugin` → `my-plugin`
- underscores, spaces, and punctuation are converted to `-`
- result is lower-case hyphen-delimited with consecutive hyphens collapsed
- Supports optional creation of:
- `skills/`
- `hooks/`
- `scripts/`
- `assets/`
- `.mcp.json`
- `.app.json`
## Marketplace workflow
- Personal-marketplace creation defaults to `~/.agents/plugins/marketplace.json`. Here,
"personal marketplace" means the marketplace whose file is at that path.
- Repo/team marketplace creation is opt-in through both `--path` and `--marketplace-path`, only
when the user specifically requests it.
- `--marketplace-name` is an exception path. Use it only when the default `personal` marketplace
name is already taken and you need to seed a different new marketplace file.
- Do not use `--marketplace-name` to rename an existing marketplace file in place. If the file
already exists, its top-level `name` must already match.
- If the user specifies a different marketplace path, treat that marketplace as needing explicit installation via `codex plugin marketplace add`.
- Prefer `scripts/read_marketplace_name.py` when you need the marketplace name from any
`marketplace.json` file. With no argument it reads the default personal marketplace; with an
explicit path it works for repo/team marketplaces too.
- In either location, the generated source path remains `./plugins/<plugin-name>`.
- Marketplace root metadata supports top-level `name` plus optional `interface.displayName`.
- Treat plugin order in `plugins[]` as render order in Codex. Append new entries unless a user explicitly asks to reorder the list.
- `displayName` belongs inside the marketplace `interface` object, not individual `plugins[]` entries.
- Each generated marketplace entry must include all of:
- `policy.installation`
- `policy.authentication`
- `category`
- Default new entries to:
- `policy.installation: "AVAILABLE"`
- `policy.authentication: "ON_INSTALL"`
- Override defaults only when the user explicitly specifies another allowed value.
- Allowed `policy.installation` values:
- `NOT_AVAILABLE`
- `AVAILABLE`
- `INSTALLED_BY_DEFAULT`
- Allowed `policy.authentication` values:
- `ON_INSTALL`
- `ON_USE`
- Treat `policy.products` as an override. Omit it unless the user explicitly requests product gating.
- The generated plugin entry shape is:
```json
{
"name": "plugin-name",
"source": {
"source": "local",
"path": "./plugins/plugin-name"
},
"policy": {
"installation": "AVAILABLE",
"authentication": "ON_INSTALL"
},
"category": "Productivity"
}
```
- Use `--force` only when intentionally replacing an existing marketplace entry for the same plugin name.
- If the target marketplace file does not exist yet, create it with top-level `"name"`, an `"interface"` object containing `"displayName"`, and a `plugins` array, then add the new entry.
- For a brand-new marketplace file, the root object should look like:
```json
{
"name": "personal",
"interface": {
"displayName": "Personal"
},
"plugins": [
{
"name": "plugin-name",
"source": {
"source": "local",
"path": "./plugins/plugin-name"
},
"policy": {
"installation": "AVAILABLE",
"authentication": "ON_INSTALL"
},
"category": "Productivity"
}
]
}
```
## Required behavior
- Outer folder name and `plugin.json` `"name"` are always the same normalized plugin name.
- Do not remove required structure; keep `.codex-plugin/plugin.json` present.
- Do not leave `[TODO: ...]` placeholders in plugin manifests.
- Keep `apps` and `mcpServers` out of `plugin.json` unless their companion files are actually created.
- Omit unsupported plugin manifest fields that validation rejects, including `hooks`.
- If creating files inside an existing plugin path, use `--force` only when overwrite is intentional.
- Preserve any existing marketplace `interface.displayName`.
- When generating marketplace entries, always write `policy.installation`, `policy.authentication`, and `category` even if their values are defaults.
- Add `policy.products` only when the user explicitly asks for that override.
- Keep marketplace `source.path` relative to the selected marketplace root as `./plugins/<plugin-name>`.
- Only use `--marketplace-name` when creating a new marketplace file whose name should not be
`personal` because that name is already taken or installed elsewhere.
- If Codex would need approval to write the marketplace file, ask for that approval before
proceeding. If the user prefers to run the write themselves, provide the exact scaffold command
and then continue from validation or subsequent plugin edits instead of leaving the workflow
vague.
- For updates to an existing local plugin during development, do not hand-edit marketplace config
or `marketplace.json`. Use the update flow documented in
`references/installing-and-updating.md` and `scripts/update_plugin_cachebuster.py`.
- Do not tell the user to run `codex plugin marketplace add` for the default personal-marketplace
flow. That command is for explicit non-default marketplace configuration, not for the standard
`~/.agents/plugins/marketplace.json` path.
- If the user provided a non-default `--marketplace-path`, make sure that marketplace is installed
before giving reinstall instructions. Use `codex plugin marketplace add <path-to-marketplace-root>`
when that explicit marketplace has not been configured yet.
- When the workflow created or updated a marketplace-backed plugin, end the final user-facing
response with a short Codex app handoff. Say `To view this in the Codex app:` and write
`View <normalized plugin name>` and `Share <normalized plugin name>` as Markdown links, not raw
URLs or code spans.
- The View deeplink uses `codex://plugins/<normalized plugin name>?marketplacePath=<absolute marketplace.json path>`.
The Share deeplink uses the same URL with `&mode=share`.
- Replace the placeholders with the real normalized plugin name and absolute `marketplace.json`
path from the scaffolded plugin. URL-encode the path segment and query value when needed.
- Do not add `pluginName` or `hostId` query parameters to these deeplinks. Codex derives both after
the user clicks the link.
- Do not emit the `View <normalized plugin name>` or `Share <normalized plugin name>` links when no marketplace entry was
created or updated.
## Reference to exact spec sample
For the exact canonical sample JSON for both plugin manifests and marketplace entries, use:
- `references/plugin-json-spec.md`
- `references/installing-and-updating.md` for update/reinstall guidance while
iterating on an existing local plugin, plus the new-thread pickup behavior after reinstall
## Validation
After editing `SKILL.md`, run:
```bash
python3 ../skill-creator/scripts/quick_validate.py .
```
Before handing back a generated plugin, run:
```bash
python3 scripts/validate_plugin.py <plugin-path>
```

6
.codex-home/skills/.system/plugin-creator/agents/openai.yaml Normal file
View file

@ -0,0 +1,6 @@
interface:
display_name: "Plugin Creator"
short_description: "Scaffold plugins and marketplace entries"
default_prompt: "Use $plugin-creator to scaffold a valid plugin in the personal marketplace, then validate it before handing it back."
icon_small: "./assets/plugin-creator-small.svg"
icon_large: "./assets/plugin-creator.png"

3
.codex-home/skills/.system/plugin-creator/assets/plugin-creator-small.svg Normal file
View file

@ -0,0 +1,3 @@
<svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" fill="currentColor" viewBox="0 0 20 20">
<path fill="#0D0D0D" d="M12.03 4.113a3.612 3.612 0 0 1 5.108 5.108l-6.292 6.29c-.324.324-.56.561-.791.752l-.235.176c-.205.14-.422.261-.65.36l-.229.093a4.136 4.136 0 0 1-.586.16l-.764.134-2.394.4c-.142.024-.294.05-.423.06-.098.007-.232.01-.378-.026l-.149-.05a1.081 1.081 0 0 1-.521-.474l-.046-.093a1.104 1.104 0 0 1-.075-.527c.01-.129.035-.28.06-.422l.398-2.394c.1-.602.162-.987.295-1.35l.093-.23c.1-.228.22-.445.36-.65l.176-.235c.19-.232.428-.467.751-.79l6.292-6.292Zm-5.35 7.232c-.35.35-.534.535-.66.688l-.11.147a2.67 2.67 0 0 0-.24.433l-.062.154c-.08.22-.124.462-.232 1.112l-.398 2.394-.001.001h.003l2.393-.399.717-.126a2.63 2.63 0 0 0 .394-.105l.154-.063a2.65 2.65 0 0 0 .433-.24l.147-.11c.153-.126.339-.31.688-.66l4.988-4.988-3.227-3.226-4.987 4.988Zm9.517-6.291a2.281 2.281 0 0 0-3.225 0l-.364.362 3.226 3.227.363-.364c.89-.89.89-2.334 0-3.225ZM4.583 1.783a.3.3 0 0 1 .294.241c.117.585.347 1.092.707 1.48.357.385.859.668 1.549.783a.3.3 0 0 1 0 .592c-.69.115-1.192.398-1.549.783-.315.34-.53.77-.657 1.265l-.05.215a.3.3 0 0 1-.588 0c-.117-.585-.347-1.092-.707-1.48-.357-.384-.859-.668-1.549-.783a.3.3 0 0 1 0-.592c.69-.115 1.192-.398 1.549-.783.36-.388.59-.895.707-1.48l.015-.05a.3.3 0 0 1 .279-.19Z"/>
</svg>
Side by side

After

Width:  |  Height:  |  Size: 1.3 KiB

BIN
.codex-home/skills/.system/plugin-creator/assets/plugin-creator.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 1.5 KiB

143
.codex-home/skills/.system/plugin-creator/references/installing-and-updating.md Normal file
View file

@ -0,0 +1,143 @@
# Updating Existing Local Plugins
Use this reference when a plugin already exists and the request is about updating the plugin during
local development.
All scripts here are specified relative to the skill root. Update the path for running the scripts
depending on your current working directory.
## When To Use This Flow
Use this flow when all of the following are true:
- the plugin already exists locally
- the marketplace entry already points at the plugin source you are editing
- the user wants Codex to see the updated plugin without manually editing marketplace files
If the user still needs the initial plugin entry or marketplace structure created, use the scaffold
flow first and only then switch to this reinstall flow.
## Update Loop
1. Update the plugin manifest to a single Codex cachebuster suffix:
```bash
python3 scripts/update_plugin_cachebuster.py \
<plugin-path>
```
Prefer the default helper behavior here. If you omit `--cachebuster`, the helper uses a UTC
timestamp down to seconds, which is the recommended path for routine local iteration.
Only use a manual cachebuster override when the user explicitly asks for one or when a workflow
outside Codex depends on a specific token:
```bash
python3 scripts/update_plugin_cachebuster.py \
<plugin-path> \
--cachebuster local-20260519-184516
```
2. For the default scaffolded flow, read the marketplace name from the personal marketplace file:
```bash
python3 scripts/read_marketplace_name.py
```
Here, "personal marketplace" means the marketplace whose file is at
`~/.agents/plugins/marketplace.json`. On Windows, use the equivalent path under the user profile.
The helper uses Python's home-directory resolution and prints the marketplace name to use when
constructing the install command.
To read the name from a different marketplace file, pass the path directly:
```bash
python3 scripts/read_marketplace_name.py --marketplace-path <path-to-marketplace.json>
```
3. Reinstall from that marketplace name:
```bash
codex plugin add <plugin-name>@<marketplace-name-from-marketplace-json>
```
The default personal marketplace is discovered implicitly from
`~/.agents/plugins/marketplace.json`. You do not need `codex plugin marketplace add` for that
path, and `codex plugin marketplace list` is not the right check for whether that default
marketplace exists.
4. If the plugin is not using the personal marketplace file, check which configured local
marketplace is actually surfacing that plugin:
```bash
codex plugin list
```
If the plugin is not in the personal marketplace file, confirm which marketplace entry points at
the plugin source you are editing and make sure that marketplace is still local. If it is a
different local marketplace, reinstall from that marketplace name instead of forcing the personal
marketplace flow. If it is not local, stop and help the user resolve the mismatch before
continuing.
5. If the plugin lives in a different confirmed local marketplace, substitute that marketplace
name:
```bash
codex plugin add <plugin-name>@<local-marketplace>
```
6. Prompt the user to use a new thread to try the updated plugin, so that Codex picks up new skills
and tools.
## Cachebuster Policy
- Preserve the existing version prefix and replace only the suffix.
- Treat the preserved prefix as everything before `+`.
- Use the format:
```text
<base-version>+codex.<cachebuster>
```
Examples:
- `0.1.0` → `0.1.0+codex.local-20260519-184516`
- `0.1.0+codex.old-token` → `0.1.0+codex.local-20260519-184516`
- `1.2.3-beta.1+codex.prev` → `1.2.3-beta.1+codex.local-20260519-184516`
- `dev-build+other-tag` → `dev-build+codex.local-20260519-184516`
Replace the existing Codex cachebuster instead of appending another one. Do not keep incrementing
numeric version components just to trigger reinstall behavior.
## Marketplace Rules
- Marketplace manipulation should happen through commands, not by hand-editing `marketplace.json`
or `config.toml` during this update/reinstall flow.
- Prefer the personal marketplace file for the default scaffolded flow.
- Read the personal marketplace name with
`python3 scripts/read_marketplace_name.py` and use the printed value when constructing
`codex plugin add <plugin-name>@<marketplace-name>`.
- For non-default marketplace files, use
`python3 scripts/read_marketplace_name.py --marketplace-path <path-to-marketplace.json>` to read
the name before constructing reinstall commands.
- Do not tell the user to run `codex plugin marketplace add` for the default personal-marketplace
flow. That marketplace is discovered implicitly by Codex.
- If the user specified a different marketplace path, make sure that marketplace is installed
before giving install or reinstall instructions. Non-default marketplace paths are not
discovered implicitly.
- Use `codex plugin list` when the plugin lives in a different configured marketplace and you need
to confirm which marketplace is surfacing that plugin.
- If a non-default local marketplace has not been configured yet, install it with
`codex plugin marketplace add <path-to-marketplace-root>` before telling the user to run
`codex plugin add <plugin-name>@<marketplace-name>`.
- If the plugin is not in the personal marketplace file, confirm that the selected marketplace is
local before telling the user to reinstall from it.
- If the selected marketplace is not local, stop and help the user resolve that mismatch rather
than pretending the normal local reinstall flow applies.
- If the plugin source is not already the source referenced by the chosen marketplace entry, stop
and fix that first. This update flow does not rewrite marketplace entries.
## After Reinstall
After reinstalling, prompt the user to start a new thread for testing. That is the safe boundary for
picking up the updated plugin and its MCP tools.

218
.codex-home/skills/.system/plugin-creator/references/plugin-json-spec.md Normal file
View file

@ -0,0 +1,218 @@
# Plugin JSON sample spec
```json
{
"name": "plugin-name",
"version": "1.2.0",
"description": "Brief plugin description",
"author": {
"name": "Author Name",
"email": "author@example.com",
"url": "https://github.com/author"
},
"homepage": "https://docs.example.com/plugin",
"repository": "https://github.com/author/plugin",
"license": "MIT",
"keywords": ["keyword1", "keyword2"],
"skills": "./skills/",
"hooks": "./hooks.json",
"mcpServers": "./.mcp.json",
"apps": "./.app.json",
"interface": {
"displayName": "Plugin Display Name",
"shortDescription": "Short description for subtitle",
"longDescription": "Long description for details page",
"developerName": "OpenAI",
"category": "Productivity",
"capabilities": ["Interactive", "Write"],
"websiteURL": "https://openai.com/",
"privacyPolicyURL": "https://openai.com/policies/row-privacy-policy/",
"termsOfServiceURL": "https://openai.com/policies/row-terms-of-use/",
"defaultPrompt": [
"Summarize my inbox and draft replies for me.",
"Find open bugs and turn them into Linear tickets.",
"Review today's meetings and flag scheduling gaps."
],
"brandColor": "#3B82F6",
"composerIcon": "./assets/icon.png",
"logo": "./assets/logo.png",
"logoDark": "./assets/logo-dark.png",
"screenshots": [
"./assets/screenshot1.png",
"./assets/screenshot2.png",
"./assets/screenshot3.png"
]
}
}
```
## Field guide
### Top-level fields
- `name` (`string`): Plugin identifier (kebab-case, no spaces). Required if `plugin.json` is provided and used as manifest name and component namespace.
- `version` (`string`): Plugin semantic version.
- `description` (`string`): Short purpose summary.
- `author` (`object`): Publisher identity.
- `name` (`string`): Author or team name.
- `email` (`string`): Contact email.
- `url` (`string`): Author/team homepage or profile URL.
- `homepage` (`string`): Documentation URL for plugin usage.
- `repository` (`string`): Source code URL.
- `license` (`string`): License identifier (for example `MIT`, `Apache-2.0`).
- `keywords` (`array` of `string`): Search/discovery tags.
- `skills` (`string`): Relative path to skill directories/files.
- `hooks` (`string`): Hook config path.
- `mcpServers` (`string` or `object`): MCP config path, or an object whose keys are MCP server names and whose values are MCP server config objects.
- `apps` (`string`): App manifest path for plugin integrations.
- `interface` (`object`): Interface/UX metadata block for plugin presentation.
`mcpServers` may be declared as a companion file path:
```json
{
"mcpServers": "./.mcp.json"
}
```
Or as an object directly in `plugin.json`:
```json
{
"mcpServers": {
"counter": {
"type": "http",
"url": "https://sample.example/counter/mcp"
}
}
}
```
### `interface` fields
- `displayName` (`string`): User-facing title shown for the plugin.
- `shortDescription` (`string`): Brief subtitle used in compact views.
- `longDescription` (`string`): Longer description used on details screens.
- `developerName` (`string`): Human-readable publisher name.
- `category` (`string`): Plugin category bucket.
- `capabilities` (`array` of `string`): Capability list from implementation.
- `websiteURL` (`string`): Public website for the plugin.
- `privacyPolicyURL` (`string`): Privacy policy URL.
- `termsOfServiceURL` (`string`): Terms of service URL.
- `defaultPrompt` (`array` of `string`): Starter prompts shown in composer/UX context.
- Include at most 3 strings. Entries after the first 3 are ignored and will not be included.
- Each string is capped at 128 characters. Longer entries are truncated.
- Prefer short starter prompts around 50 characters so they scan well in the UI.
- `brandColor` (`string`): Theme color for the plugin card.
- `composerIcon` (`string`): Path to icon asset.
- `logo` (`string`): Path to logo asset.
- `logoDark` (`string`): Optional path to the logo asset used in dark mode.
- `screenshots` (`array` of `string`): List of screenshot asset paths.
- Screenshot entries must be PNG filenames and stored under `./assets/`.
- Keep file paths relative to plugin root.
### Path conventions and defaults
- Path values should be relative and begin with `./`.
- `skills`, `hooks`, and string-valued `mcpServers` are supplemented on top of default component discovery; they do not replace defaults.
- Custom path values must follow the plugin root convention and naming/namespacing rules.
- This repo’s scaffold writes `.codex-plugin/plugin.json`; treat that as the manifest location this skill generates.
# Marketplace JSON sample spec
`marketplace.json` depends on where the plugin should live. New plugin creation defaults to the
personal marketplace unless the caller explicitly requests a repo-local destination:
- Personal plugin: `~/.agents/plugins/marketplace.json`
- Repo/team plugin: `<repo-root>/.agents/plugins/marketplace.json`
```json
{
"name": "openai-curated",
"interface": {
"displayName": "ChatGPT Official"
},
"plugins": [
{
"name": "linear",
"source": {
"source": "local",
"path": "./plugins/linear"
},
"policy": {
"installation": "AVAILABLE",
"authentication": "ON_INSTALL"
},
"category": "Productivity"
}
]
}
```
## Marketplace field guide
### Top-level fields
- `name` (`string`): Marketplace identifier or catalog name.
- `interface` (`object`, optional): Marketplace presentation metadata.
- `plugins` (`array`): Ordered plugin entries. This order determines how Codex renders plugins.
### `interface` fields
- `displayName` (`string`, optional): User-facing marketplace title.
### Plugin entry fields
- `name` (`string`): Plugin identifier. Match the plugin folder name and `plugin.json` `name`.
- `source` (`object`): Plugin source descriptor.
- `source` (`string`): Use `local` for this repo workflow.
- `path` (`string`): Relative plugin path based on the marketplace root.
- Personal plugin in `~/.agents/plugins/marketplace.json`: `./plugins/<plugin-name>`
- Repo/team plugin: `./plugins/<plugin-name>`
- The same relative path convention is used for both personal and repo/team marketplaces.
- Example: with `~/.agents/plugins/marketplace.json`, `./plugins/<plugin-name>` resolves to
`~/plugins/<plugin-name>`.
- `policy` (`object`): Marketplace policy block. Always include it.
- `installation` (`string`): Availability policy.
- Allowed values: `NOT_AVAILABLE`, `AVAILABLE`, `INSTALLED_BY_DEFAULT`
- Default for new entries: `AVAILABLE`
- `authentication` (`string`): Authentication timing policy.
- Allowed values: `ON_INSTALL`, `ON_USE`
- Default for new entries: `ON_INSTALL`
- `products` (`array` of `string`, optional): Product override for this plugin entry. Omit it unless product gating is explicitly requested.
- `category` (`string`): Display category bucket. Always include it.
### Marketplace generation rules
- `displayName` belongs under the top-level `interface` object, not individual plugin entries.
- When creating a new marketplace file from scratch, seed `interface.displayName` alongside top-level `name`.
- Always include `policy.installation`, `policy.authentication`, and `category` on every generated or updated plugin entry.
- Treat `policy.products` as an override and omit it unless explicitly requested.
- Append new entries unless the user explicitly requests reordering.
- Replace an existing entry for the same plugin only when overwrite is intentional.
- Default new plugin creation to the personal marketplace.
- Use a repo/team marketplace only when the user specifically requests that destination.
- Only override the marketplace `name` when the default `personal` name is already taken or
installed and you need to seed a different new marketplace file.
- Choose marketplace location to match the selected destination:
- Personal plugin: `~/.agents/plugins/marketplace.json`
- Repo/team plugin: `<repo-root>/.agents/plugins/marketplace.json`
### Plugin validation notes
- The validator mirrors the workspace plugin ingestion schema so generated plugins follow the same
manifest contract from the start.
- Plugin manifests must include real values for `name`, `version`, `description`,
`author.name`, and the required `interface` fields.
- `version` must use strict semver.
- `websiteURL`, `privacyPolicyURL`, and `termsOfServiceURL` must be absolute `https://` URLs when
present.
- `composerIcon`, `logo`, `logoDark`, and `screenshots` must point to real files inside the plugin archive when
present.
- `apps` should appear in `plugin.json` only when `.app.json` actually exists.
- `mcpServers` may point to `.mcp.json` or contain the MCP server object directly in
`plugin.json`.
- Validation rejects unsupported manifest fields such as `hooks`, so the scaffold keeps them out of
generated manifests.
- Run `scripts/validate_plugin.py <plugin-path>` before handing back a generated plugin. It adds one
intentional preflight check that rejects leftover `[TODO: ...]` placeholders.

324
.codex-home/skills/.system/plugin-creator/scripts/create_basic_plugin.py Normal file
View file

@ -0,0 +1,324 @@
#!/usr/bin/env python3
"""Scaffold a plugin directory and optionally update marketplace.json."""
from __future__ import annotations
import argparse
import json
import re
from pathlib import Path
from typing import Any
MAX_PLUGIN_NAME_LENGTH = 64
DEFAULT_INSTALL_POLICY = "AVAILABLE"
DEFAULT_AUTH_POLICY = "ON_INSTALL"
DEFAULT_CATEGORY = "Productivity"
DEFAULT_MARKETPLACE_NAME = "personal"
VALID_INSTALL_POLICIES = {"NOT_AVAILABLE", "AVAILABLE", "INSTALLED_BY_DEFAULT"}
VALID_AUTH_POLICIES = {"ON_INSTALL", "ON_USE"}
DEFAULT_PLUGIN_PARENT = Path.home() / "plugins"
DEFAULT_MARKETPLACE_PATH = Path.home() / ".agents" / "plugins" / "marketplace.json"
def normalize_plugin_name(plugin_name: str) -> str:
"""Normalize a plugin name to lowercase hyphen-case."""
normalized = plugin_name.strip().lower()
normalized = re.sub(r"[^a-z0-9]+", "-", normalized)
normalized = normalized.strip("-")
normalized = re.sub(r"-{2,}", "-", normalized)
return normalized
def validate_plugin_name(plugin_name: str) -> None:
if not plugin_name:
raise ValueError("Plugin name must include at least one letter or digit.")
if len(plugin_name) > MAX_PLUGIN_NAME_LENGTH:
raise ValueError(
f"Plugin name '{plugin_name}' is too long ({len(plugin_name)} characters). "
f"Maximum is {MAX_PLUGIN_NAME_LENGTH} characters."
)
def validate_marketplace_name(marketplace_name: str) -> None:
if not marketplace_name:
raise ValueError("Marketplace name must include at least one letter or digit.")
if re.fullmatch(r"[A-Za-z0-9_-]+", marketplace_name) is None:
raise ValueError(
"Marketplace name may only contain ASCII letters, digits, `_`, and `-`."
)
def display_name_from_plugin_name(plugin_name: str) -> str:
return " ".join(part.capitalize() for part in re.split(r"[-_]+", plugin_name))
def build_plugin_json(plugin_name: str, *, with_mcp: bool, with_apps: bool) -> dict[str, Any]:
display_name = display_name_from_plugin_name(plugin_name)
payload: dict[str, Any] = {
"name": plugin_name,
"version": "0.1.0",
"description": f"{display_name} plugin",
"author": {
"name": "Local developer",
},
"skills": "./skills/",
"interface": {
"displayName": display_name,
"shortDescription": f"Use {display_name} in Codex.",
"longDescription": f"{display_name} adds a local Codex plugin scaffold.",
"developerName": "Local developer",
"category": DEFAULT_CATEGORY,
"capabilities": [],
"defaultPrompt": f"Help me use {display_name}.",
},
}
if with_mcp:
payload["mcpServers"] = "./.mcp.json"
if with_apps:
payload["apps"] = "./.app.json"
return payload
def build_marketplace_entry(
plugin_name: str,
install_policy: str,
auth_policy: str,
category: str,
) -> dict[str, Any]:
return {
"name": plugin_name,
"source": {
"source": "local",
"path": f"./plugins/{plugin_name}",
},
"policy": {
"installation": install_policy,
"authentication": auth_policy,
},
"category": category,
}
def load_json(path: Path) -> dict[str, Any]:
with path.open() as handle:
return json.load(handle)
def build_default_marketplace(marketplace_name: str) -> dict[str, Any]:
return {
"name": marketplace_name,
"interface": {
"displayName": display_name_from_plugin_name(marketplace_name),
},
"plugins": [],
}
def validate_marketplace_interface(payload: dict[str, Any]) -> None:
interface = payload.get("interface")
if interface is not None and not isinstance(interface, dict):
raise ValueError("marketplace.json field 'interface' must be an object.")
def update_marketplace_json(
marketplace_path: Path,
marketplace_name: str | None,
plugin_name: str,
install_policy: str,
auth_policy: str,
category: str,
force: bool,
) -> None:
if marketplace_path.exists():
payload = load_json(marketplace_path)
else:
payload = build_default_marketplace(marketplace_name or DEFAULT_MARKETPLACE_NAME)
if not isinstance(payload, dict):
raise ValueError(f"{marketplace_path} must contain a JSON object.")
validate_marketplace_interface(payload)
existing_marketplace_name = payload.get("name")
if marketplace_name is not None:
if not isinstance(existing_marketplace_name, str) or not existing_marketplace_name.strip():
raise ValueError(f"{marketplace_path} must contain a non-empty string 'name'.")
if existing_marketplace_name != marketplace_name:
raise ValueError(
f"{marketplace_path} already uses marketplace name "
f"'{existing_marketplace_name}'. Create a new marketplace file to use "
f"'{marketplace_name}' instead."
)
plugins = payload.setdefault("plugins", [])
if not isinstance(plugins, list):
raise ValueError(f"{marketplace_path} field 'plugins' must be an array.")
new_entry = build_marketplace_entry(plugin_name, install_policy, auth_policy, category)
for index, entry in enumerate(plugins):
if isinstance(entry, dict) and entry.get("name") == plugin_name:
if not force:
raise FileExistsError(
f"Marketplace entry '{plugin_name}' already exists in {marketplace_path}. "
"Use --force to overwrite that entry."
)
plugins[index] = new_entry
break
else:
plugins.append(new_entry)
write_json(marketplace_path, payload, force=True)
def write_json(path: Path, data: dict, force: bool) -> None:
if path.exists() and not force:
raise FileExistsError(f"{path} already exists. Use --force to overwrite.")
path.parent.mkdir(parents=True, exist_ok=True)
with path.open("w") as handle:
json.dump(data, handle, indent=2)
handle.write("\n")
def create_stub_file(path: Path, payload: dict, force: bool) -> None:
if path.exists() and not force:
return
path.parent.mkdir(parents=True, exist_ok=True)
with path.open("w") as handle:
json.dump(payload, handle, indent=2)
handle.write("\n")
def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser(
description="Create a plugin skeleton with a validation-ready plugin.json."
)
parser.add_argument("plugin_name")
parser.add_argument(
"--path",
default=str(DEFAULT_PLUGIN_PARENT),
help=(
"Parent directory for plugin creation (defaults to <home>/plugins). "
"Pass an explicit repo path only when a repo/team plugin is intended."
),
)
parser.add_argument("--with-skills", action="store_true", help="Create skills/ directory")
parser.add_argument("--with-hooks", action="store_true", help="Create hooks/ directory")
parser.add_argument("--with-scripts", action="store_true", help="Create scripts/ directory")
parser.add_argument("--with-assets", action="store_true", help="Create assets/ directory")
parser.add_argument("--with-mcp", action="store_true", help="Create .mcp.json placeholder")
parser.add_argument("--with-apps", action="store_true", help="Create .app.json placeholder")
parser.add_argument(
"--with-marketplace",
action="store_true",
help=(
"Create or update <home>/.agents/plugins/marketplace.json by default. "
"Marketplace entries always point to ./plugins/<plugin-name> relative to the "
"marketplace root."
),
)
parser.add_argument(
"--marketplace-path",
default=str(DEFAULT_MARKETPLACE_PATH),
help=(
"Path to marketplace.json (defaults to <home>/.agents/plugins/marketplace.json). "
"Pass a repo-rooted marketplace path only when a repo/team plugin is intended."
),
)
parser.add_argument(
"--marketplace-name",
help=(
"Marketplace name to seed into a new marketplace.json. Use this only when the default "
"'personal' marketplace name is already taken and you need a different new marketplace."
),
)
parser.add_argument(
"--install-policy",
default=DEFAULT_INSTALL_POLICY,
choices=sorted(VALID_INSTALL_POLICIES),
help="Marketplace policy.installation value",
)
parser.add_argument(
"--auth-policy",
default=DEFAULT_AUTH_POLICY,
choices=sorted(VALID_AUTH_POLICIES),
help="Marketplace policy.authentication value",
)
parser.add_argument(
"--category",
default=DEFAULT_CATEGORY,
help="Marketplace category value",
)
parser.add_argument("--force", action="store_true", help="Overwrite existing files")
return parser.parse_args()
def main() -> None:
args = parse_args()
raw_plugin_name = args.plugin_name
plugin_name = normalize_plugin_name(raw_plugin_name)
if plugin_name != raw_plugin_name:
print(f"Note: Normalized plugin name from '{raw_plugin_name}' to '{plugin_name}'.")
validate_plugin_name(plugin_name)
marketplace_name = None
if args.marketplace_name is not None:
marketplace_name = args.marketplace_name.strip()
validate_marketplace_name(marketplace_name)
plugin_root = (Path(args.path).expanduser().resolve() / plugin_name)
plugin_root.mkdir(parents=True, exist_ok=True)
plugin_json_path = plugin_root / ".codex-plugin" / "plugin.json"
write_json(
plugin_json_path,
build_plugin_json(plugin_name, with_mcp=args.with_mcp, with_apps=args.with_apps),
args.force,
)
optional_directories = {
"skills": args.with_skills,
"hooks": args.with_hooks,
"scripts": args.with_scripts,
"assets": args.with_assets,
}
for folder, enabled in optional_directories.items():
if enabled:
(plugin_root / folder).mkdir(parents=True, exist_ok=True)
if args.with_mcp:
create_stub_file(
plugin_root / ".mcp.json",
{"mcpServers": {}},
args.force,
)
if args.with_apps:
create_stub_file(
plugin_root / ".app.json",
{
"apps": {},
},
args.force,
)
if args.with_marketplace:
marketplace_path = Path(args.marketplace_path).expanduser().resolve()
update_marketplace_json(
marketplace_path,
marketplace_name,
plugin_name,
args.install_policy,
args.auth_policy,
args.category,
args.force,
)
print(f"Created plugin scaffold: {plugin_root}")
print(f"plugin manifest: {plugin_json_path}")
if args.with_marketplace:
print(f"marketplace manifest: {marketplace_path}")
if __name__ == "__main__":
main()

48
.codex-home/skills/.system/plugin-creator/scripts/read_marketplace_name.py Normal file
View file

@ -0,0 +1,48 @@
#!/usr/bin/env python3
"""Print the top-level marketplace name from any marketplace.json file."""
from __future__ import annotations
import argparse
import json
import sys
from pathlib import Path
def default_marketplace_path() -> Path:
return Path.home() / ".agents" / "plugins" / "marketplace.json"
def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser(
description=(
"Print the top-level marketplace name from marketplace.json. Defaults to the personal "
"marketplace path under the current home directory."
)
)
parser.add_argument(
"--marketplace-path",
default=str(default_marketplace_path()),
help="Path to marketplace.json",
)
return parser.parse_args()
def main() -> None:
args = parse_args()
marketplace_path = Path(args.marketplace_path).expanduser().resolve()
payload = json.loads(marketplace_path.read_text(encoding="utf-8"))
if not isinstance(payload, dict):
raise ValueError(f"{marketplace_path} must contain a JSON object.")
name = payload.get("name")
if not isinstance(name, str) or not name.strip():
raise ValueError(f"{marketplace_path} must contain a non-empty string 'name'.")
print(name.strip())
if __name__ == "__main__":
try:
main()
except Exception as err: # noqa: BLE001 - CLI should surface a single clear message.
print(str(err), file=sys.stderr)
raise SystemExit(1) from err

78
.codex-home/skills/.system/plugin-creator/scripts/update_plugin_cachebuster.py Normal file
View file

@ -0,0 +1,78 @@
#!/usr/bin/env python3
"""Rewrite a local plugin version to a single Codex cachebuster suffix."""
from __future__ import annotations
import argparse
import json
import re
import sys
from datetime import datetime, timezone
from pathlib import Path
CACHEBUSTER_PREFIX = "codex"
def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser(
description=(
"Rewrite a local plugin's version so it preserves everything before '+' and uses "
"a single +codex.<cachebuster> suffix."
)
)
parser.add_argument("plugin_path", help="Path to the plugin root directory")
parser.add_argument(
"--cachebuster",
help="Optional cachebuster token to embed in the plugin version",
)
return parser.parse_args()
def main() -> None:
args = parse_args()
plugin_root = Path(args.plugin_path).expanduser().resolve()
manifest_path = plugin_root / ".codex-plugin" / "plugin.json"
manifest = load_manifest(manifest_path)
version = manifest.get("version")
if not isinstance(version, str) or not version.strip():
raise ValueError(f"{manifest_path} must contain a non-empty string 'version'.")
cachebuster = sanitize_cachebuster(args.cachebuster or default_cachebuster())
next_version = with_cachebuster(version, cachebuster)
manifest["version"] = next_version
manifest_path.write_text(json.dumps(manifest, indent=2) + "\n", encoding="utf-8")
print(f"Updated plugin version: {version} -> {next_version}")
def load_manifest(manifest_path: Path) -> dict[str, object]:
if not manifest_path.is_file():
raise FileNotFoundError(f"missing manifest: {manifest_path}")
payload = json.loads(manifest_path.read_text(encoding="utf-8"))
if not isinstance(payload, dict):
raise ValueError(f"{manifest_path} must contain a JSON object.")
return payload
def sanitize_cachebuster(value: str) -> str:
sanitized = re.sub(r"[^a-z0-9-]+", "-", value.strip().lower())
sanitized = re.sub(r"-{2,}", "-", sanitized).strip("-")
if not sanitized:
raise ValueError("Cachebuster must contain at least one letter or digit.")
return sanitized
def default_cachebuster() -> str:
return datetime.now(timezone.utc).strftime("%Y%m%d%H%M%S")
def with_cachebuster(version: str, cachebuster: str) -> str:
version_prefix = version.split("+", 1)[0]
return f"{version_prefix}+{CACHEBUSTER_PREFIX}.{cachebuster}"
if __name__ == "__main__":
try:
main()
except Exception as err: # noqa: BLE001 - CLI should surface a single clear message.
print(str(err), file=sys.stderr)
raise SystemExit(1) from err

629
.codex-home/skills/.system/plugin-creator/scripts/validate_plugin.py Normal file
View file

@ -0,0 +1,629 @@
#!/usr/bin/env python3
"""Validate a generated plugin against the plugin ingestion contract."""
from __future__ import annotations
import argparse
import json
import re
from pathlib import Path, PurePosixPath
from typing import Any
from urllib.parse import urlparse
import yaml
TODO_MARKER = "[TODO:"
SEMVER_RE = re.compile(
r"^(0|[1-9]\d*)\."
r"(0|[1-9]\d*)\."
r"(0|[1-9]\d*)"
r"(?:-(?:0|[1-9]\d*|\d*[A-Za-z-][0-9A-Za-z-]*)(?:\."
r"(?:0|[1-9]\d*|\d*[A-Za-z-][0-9A-Za-z-]*))*)?"
r"(?:\+[0-9A-Za-z-]+(?:\.[0-9A-Za-z-]+)*)?$"
)
HEX_COLOR_RE = re.compile(r"^#[0-9A-F]{6}$", re.IGNORECASE)
def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser(description="Validate a local Codex plugin.")
parser.add_argument("plugin_path", help="Path to the plugin root directory")
return parser.parse_args()
def main() -> None:
args = parse_args()
plugin_root = Path(args.plugin_path).expanduser().resolve()
errors = validate_plugin(plugin_root)
if errors:
print("Plugin validation failed:")
for error in errors:
print(f"- {error}")
raise SystemExit(1)
print(f"Plugin validation passed: {plugin_root}")
def validate_plugin(plugin_root: Path) -> list[str]:
errors: list[str] = []
manifest_path = plugin_root / ".codex-plugin" / "plugin.json"
manifest = load_json_object(manifest_path, errors)
if manifest is None:
return errors
reject_todo_markers(manifest, "$", errors)
validate_manifest_shape(plugin_root, manifest, errors)
return errors
def load_json_object(path: Path, errors: list[str]) -> dict[str, Any] | None:
if not path.is_file():
errors.append("missing `.codex-plugin/plugin.json`")
return None
try:
payload = json.loads(path.read_text(encoding="utf-8"))
except OSError:
errors.append("unable to read `.codex-plugin/plugin.json`")
return None
except json.JSONDecodeError:
errors.append("`.codex-plugin/plugin.json` must be valid JSON")
return None
if not isinstance(payload, dict):
errors.append("`.codex-plugin/plugin.json` must contain a JSON object")
return None
return payload
def reject_todo_markers(value: Any, path: str, errors: list[str]) -> None:
if isinstance(value, str):
if TODO_MARKER in value:
errors.append(f"{path} still contains a `[TODO: ...]` placeholder")
return
if isinstance(value, list):
for index, item in enumerate(value):
reject_todo_markers(item, f"{path}[{index}]", errors)
return
if isinstance(value, dict):
for key, item in value.items():
reject_todo_markers(item, f"{path}.{key}", errors)
def validate_manifest_shape(
plugin_root: Path,
manifest: dict[str, Any],
errors: list[str],
) -> None:
allowed_keys = {
"id",
"name",
"version",
"description",
"skills",
"apps",
"mcpServers",
"interface",
"author",
"homepage",
"repository",
"license",
"keywords",
}
for key in sorted(set(manifest) - allowed_keys):
errors.append(f"plugin.json field `{key}` is not accepted by plugin validation")
validate_optional_non_empty_string(manifest, "id", errors)
require_non_empty_string(manifest, "name", errors)
version = require_non_empty_string(manifest, "version", errors)
if version is not None and SEMVER_RE.fullmatch(version) is None:
errors.append("plugin.json field `version` must be strict semver")
require_non_empty_string(manifest, "description", errors)
author = require_object(manifest, "author", errors)
if author is not None:
reject_unknown_fields(author, {"name", "email", "url"}, "author", errors)
require_non_empty_string(author, "name", errors, prefix="author")
validate_optional_non_empty_string(author, "email", errors, prefix="author")
validate_optional_https_url(author, "url", errors, prefix="author")
validate_optional_contract_path(manifest, "skills", "skills", errors)
validate_optional_contract_path(manifest, "apps", ".app.json", errors)
validate_manifest_mcp_servers(plugin_root, manifest, errors)
if manifest.get("apps") is not None:
validate_app_manifest(
plugin_root / ".app.json",
errors,
)
validate_skill_manifests(plugin_root, errors)
interface = require_object(manifest, "interface", errors)
if interface is None:
return
reject_unknown_fields(
interface,
{
"displayName",
"shortDescription",
"longDescription",
"developerName",
"category",
"capabilities",
"websiteURL",
"privacyPolicyURL",
"termsOfServiceURL",
"brandColor",
"composerIcon",
"logo",
"logoDark",
"screenshots",
"defaultPrompt",
"default_prompt",
},
"interface",
errors,
)
for field in (
"displayName",
"shortDescription",
"longDescription",
"developerName",
"category",
):
require_non_empty_string(interface, field, errors, prefix="interface")
if "defaultPrompt" not in interface and "default_prompt" not in interface:
errors.append(
"plugin.json field `interface.defaultPrompt` or `interface.default_prompt` is required"
)
capabilities = interface.get("capabilities")
if not isinstance(capabilities, list) or not all(
isinstance(value, str) and value.strip() for value in capabilities
):
errors.append("plugin.json field `interface.capabilities` must be an array of strings")
for field in ("websiteURL", "privacyPolicyURL", "termsOfServiceURL"):
validate_optional_https_url(interface, field, errors, prefix="interface")
brand_color = interface.get("brandColor")
if brand_color is not None and (
not isinstance(brand_color, str) or HEX_COLOR_RE.fullmatch(brand_color) is None
):
errors.append("plugin.json field `interface.brandColor` must use `#RRGGBB`")
for field in ("composerIcon", "logo", "logoDark"):
validate_optional_asset_path(plugin_root, plugin_root, interface, field, errors)
screenshots = interface.get("screenshots", [])
if not isinstance(screenshots, list):
errors.append("plugin.json field `interface.screenshots` must be an array")
else:
for index, raw_path in enumerate(screenshots):
validate_asset_path(
plugin_root,
plugin_root,
raw_path,
f"interface.screenshots[{index}]",
errors,
)
def require_object(
payload: dict[str, Any],
key: str,
errors: list[str],
) -> dict[str, Any] | None:
value = payload.get(key)
if not isinstance(value, dict):
errors.append(f"plugin.json field `{key}` must be an object")
return None
return value
def require_non_empty_string(
payload: dict[str, Any],
key: str,
errors: list[str],
*,
prefix: str | None = None,
) -> str | None:
value = payload.get(key)
field = f"{prefix}.{key}" if prefix is not None else key
if not isinstance(value, str) or not value.strip():
errors.append(f"plugin.json field `{field}` must be a non-empty string")
return None
return value
def validate_optional_non_empty_string(
payload: dict[str, Any],
key: str,
errors: list[str],
*,
prefix: str | None = None,
) -> None:
value = payload.get(key)
if value is None:
return
field = f"{prefix}.{key}" if prefix is not None else key
if not isinstance(value, str) or not value.strip():
errors.append(f"plugin.json field `{field}` must be a non-empty string")
def reject_unknown_fields(
payload: dict[str, Any],
allowed_keys: set[str],
prefix: str,
errors: list[str],
) -> None:
for key in sorted(set(payload) - allowed_keys):
errors.append(f"plugin.json field `{prefix}.{key}` is not accepted by plugin validation")
def validate_optional_https_url(
payload: dict[str, Any],
key: str,
errors: list[str],
*,
prefix: str,
) -> None:
value = payload.get(key)
if value is None:
return
parsed = urlparse(value) if isinstance(value, str) else None
if parsed is None or parsed.scheme != "https" or not parsed.netloc:
errors.append(f"plugin.json field `{prefix}.{key}` must be an absolute `https://` URL")
def validate_optional_contract_path(
payload: dict[str, Any],
key: str,
expected: str,
errors: list[str],
) -> None:
value = payload.get(key)
if value is None:
return
normalized = normalize_contract_path(value) if isinstance(value, str) else None
if normalized != expected:
errors.append(f"plugin.json field `{key}` must resolve to `{expected}`")
def validate_manifest_mcp_servers(
plugin_root: Path,
manifest: dict[str, Any],
errors: list[str],
) -> None:
value = manifest.get("mcpServers")
if value is None:
return
if isinstance(value, str):
validate_optional_contract_path(manifest, "mcpServers", ".mcp.json", errors)
validate_mcp_manifest(
plugin_root / ".mcp.json",
errors,
)
return
if isinstance(value, dict):
validate_mcp_server_entries(
value,
"plugin.json field `mcpServers`",
"plugin.json field `mcpServers`",
errors,
)
return
errors.append("plugin.json field `mcpServers` must be a string path or object")
def normalize_contract_path(raw_path: str) -> str | None:
path = Path(raw_path)
if path.is_absolute():
return None
normalized = path.as_posix().rstrip("/")
return normalized or None
def validate_app_manifest(path: Path, errors: list[str]) -> None:
payload = load_companion_json_object(path, "`.app.json`", errors)
if payload is None:
return
reject_companion_unknown_fields(payload, {"apps"}, "`.app.json`", errors)
apps = payload.get("apps")
if not isinstance(apps, dict):
errors.append("`.app.json` field `apps` must be an object")
return
for key, value in apps.items():
if not isinstance(value, dict):
errors.append(f"`.app.json` app `{key}` must be an object")
continue
reject_companion_unknown_fields(
value, {"id", "category"}, f"`.app.json` app `{key}`", errors
)
app_id = value.get("id")
if not isinstance(app_id, str) or not app_id.strip():
errors.append(f"`.app.json` app `{key}` field `id` must be a non-empty string")
category = value.get("category")
if category is not None and (not isinstance(category, str) or not category.strip()):
errors.append(
f"`.app.json` app `{key}` field `category` must be a non-empty string"
)
def validate_mcp_manifest(path: Path, errors: list[str]) -> None:
payload = load_companion_json_object(path, "`.mcp.json`", errors)
if payload is None:
return
reject_companion_unknown_fields(payload, {"mcpServers"}, "`.mcp.json`", errors)
servers = payload.get("mcpServers")
validate_mcp_server_entries(
servers,
"`.mcp.json`",
"`.mcp.json` field `mcpServers`",
errors,
)
def validate_mcp_server_entries(
servers: Any,
source_label: str,
field_label: str,
errors: list[str],
) -> None:
if not isinstance(servers, dict):
errors.append(f"{field_label} must be an object")
return
for key, value in servers.items():
if not isinstance(key, str) or not key.strip():
errors.append(f"{source_label} server names must be non-empty strings")
if not isinstance(value, dict):
errors.append(f"{source_label} server `{key}` must be an object")
def load_companion_json_object(
path: Path,
label: str,
errors: list[str],
) -> dict[str, Any] | None:
if not path.is_file():
errors.append(f"{label} is required when its plugin.json field is present")
return None
try:
payload = json.loads(path.read_text(encoding="utf-8"))
except (OSError, json.JSONDecodeError):
errors.append(f"{label} must contain valid JSON")
return None
if not isinstance(payload, dict):
errors.append(f"{label} must contain a JSON object")
return None
return payload
def reject_companion_unknown_fields(
payload: dict[str, Any],
allowed_keys: set[str],
prefix: str,
errors: list[str],
) -> None:
for key in sorted(set(payload) - allowed_keys):
errors.append(f"{prefix} field `{key}` is not accepted by plugin validation")
def validate_skill_manifests(plugin_root: Path, errors: list[str]) -> None:
skills_root = plugin_root / "skills"
if not skills_root.is_dir():
return
for skill_root in sorted(skills_root.iterdir(), key=lambda path: path.name):
if skill_root.name.startswith(".") or not skill_root.is_dir():
continue
validate_skill_manifest(skill_root, errors)
def validate_skill_manifest(skill_root: Path, errors: list[str]) -> None:
skill_md_path = skill_root / "SKILL.md"
if not skill_md_path.is_file():
errors.append(f"skill `{skill_root.name}` is missing `SKILL.md`")
return
try:
contents = skill_md_path.read_text(encoding="utf-8")
except OSError:
errors.append(f"unable to read skill `{skill_root.name}`")
return
if not contents.startswith("---\n"):
errors.append(f"skill `{skill_root.name}` must start with YAML frontmatter")
return
frontmatter_end = contents.find("\n---", 4)
if frontmatter_end == -1:
errors.append(f"skill `{skill_root.name}` frontmatter is not closed")
return
try:
frontmatter = yaml.safe_load(contents[4:frontmatter_end])
except yaml.YAMLError:
errors.append(f"skill `{skill_root.name}` frontmatter must be valid YAML")
return
if not isinstance(frontmatter, dict):
errors.append(f"skill `{skill_root.name}` frontmatter must be an object")
return
skill_name = frontmatter.get("name")
if not isinstance(skill_name, str) or not skill_name.strip():
errors.append(f"skill `{skill_root.name}` frontmatter field `name` must be non-empty")
description = frontmatter.get("description")
if not isinstance(description, str) or not description.strip():
errors.append(
f"skill `{skill_root.name}` frontmatter field `description` must be non-empty"
)
disable_model_invocation = frontmatter.get("disable-model-invocation")
if disable_model_invocation is None:
disable_model_invocation = frontmatter.get("disable_model_invocation")
if disable_model_invocation not in (None, False):
errors.append(
f"skill `{skill_root.name}` frontmatter field `disable-model-invocation` must be false"
)
agent_yaml_path = skill_root / "agents" / "openai.yaml"
if agent_yaml_path.is_file():
validate_skill_agent_manifest(
plugin_root=skill_root.parent.parent,
skill_root=skill_root,
agent_yaml_path=agent_yaml_path,
errors=errors,
)
def validate_skill_agent_manifest(
*,
plugin_root: Path,
skill_root: Path,
agent_yaml_path: Path,
errors: list[str],
) -> None:
try:
payload = yaml.safe_load(agent_yaml_path.read_text(encoding="utf-8"))
except OSError:
errors.append(f"unable to read skill `{skill_root.name}` agent YAML")
return
except yaml.YAMLError:
errors.append(f"skill `{skill_root.name}` agent YAML must be valid YAML")
return
if not isinstance(payload, dict):
errors.append(f"skill `{skill_root.name}` agent YAML must be an object")
return
reject_skill_agent_unknown_fields(
payload,
{"interface", "policy", "dependencies"},
skill_root,
errors,
)
interface = payload.get("interface")
if not isinstance(interface, dict):
errors.append(f"skill `{skill_root.name}` agent field `interface` must be an object")
return
reject_skill_agent_unknown_fields(
interface,
{
"display_name",
"short_description",
"icon_small",
"icon_large",
"brand_color",
"default_prompt",
},
skill_root,
errors,
prefix="interface",
)
for field in ("display_name", "short_description"):
value = interface.get(field)
if not isinstance(value, str) or not value.strip():
errors.append(
f"skill `{skill_root.name}` agent field `interface.{field}` must be non-empty"
)
for field in ("icon_small", "icon_large"):
validate_optional_asset_path(
skill_root,
plugin_root,
interface,
field,
errors,
prefix=f"skill `{skill_root.name}` agent field `interface",
)
brand_color = interface.get("brand_color")
if brand_color is not None and (
not isinstance(brand_color, str) or HEX_COLOR_RE.fullmatch(brand_color) is None
):
errors.append(
f"skill `{skill_root.name}` agent field `interface.brand_color` must use `#RRGGBB`"
)
default_prompt = interface.get("default_prompt")
if default_prompt is not None and (
not isinstance(default_prompt, str) or not default_prompt.strip()
):
errors.append(
f"skill `{skill_root.name}` agent field `interface.default_prompt` must be non-empty"
)
policy = payload.get("policy")
if policy is not None:
if not isinstance(policy, dict):
errors.append(f"skill `{skill_root.name}` agent field `policy` must be an object")
else:
reject_skill_agent_unknown_fields(
policy,
{"allow_implicit_invocation"},
skill_root,
errors,
prefix="policy",
)
allow_implicit_invocation = policy.get("allow_implicit_invocation")
if allow_implicit_invocation is not None and not isinstance(
allow_implicit_invocation,
bool,
):
errors.append(
f"skill `{skill_root.name}` agent field "
"`policy.allow_implicit_invocation` must be a boolean"
)
dependencies = payload.get("dependencies")
if dependencies is not None:
if not isinstance(dependencies, dict):
errors.append(
f"skill `{skill_root.name}` agent field `dependencies` must be an object"
)
else:
reject_skill_agent_unknown_fields(
dependencies,
{"tools"},
skill_root,
errors,
prefix="dependencies",
)
def reject_skill_agent_unknown_fields(
payload: dict[str, Any],
allowed_keys: set[str],
skill_root: Path,
errors: list[str],
*,
prefix: str | None = None,
) -> None:
for key in sorted(set(payload) - allowed_keys):
field = f"{prefix}.{key}" if prefix is not None else key
errors.append(
f"skill `{skill_root.name}` agent field `{field}` is not accepted by plugin validation"
)
def validate_optional_asset_path(
base_dir: Path,
allowed_root: Path,
payload: dict[str, Any],
key: str,
errors: list[str],
*,
prefix: str = "interface",
) -> None:
raw_path = payload.get(key)
if raw_path is None:
return
validate_asset_path(base_dir, allowed_root, raw_path, f"{prefix}.{key}", errors)
def validate_asset_path(
base_dir: Path,
allowed_root: Path,
raw_path: Any,
field: str,
errors: list[str],
) -> None:
label = field if field.startswith("skill `") else f"plugin.json field `{field}`"
if not isinstance(raw_path, str) or not raw_path.strip():
errors.append(f"{label} must be a non-empty relative path")
return
candidate = PurePosixPath(raw_path.replace("\\", "/"))
if candidate.is_absolute() or any(part in {"", ".", ".."} for part in candidate.parts):
errors.append(f"{label} must stay inside the plugin archive")
return
resolved_path = (base_dir / candidate.as_posix()).resolve()
if not resolved_path.is_relative_to(allowed_root.resolve()):
errors.append(f"{label} must stay inside the plugin archive")
return
if not resolved_path.is_file():
errors.append(f"{label} points to a missing file")
if __name__ == "__main__":
main()

57
.codex-home/skills/.system/review-agent/SKILL.md Normal file
View file

@ -0,0 +1,57 @@
---
name: review-agent
description: Perform a read-only, defect-first review of a specified code change and return every actionable finding. Use when another agent delegates review of uncommitted changes, a base-branch diff, a commit, or custom review instructions.
---
# Review Agent
Inspect the requested target directly and return every finding that the author would likely fix.
Do not modify files, create commits, push branches, post review comments, or delegate the review
to another agent.
## Review the change
1. Read the applicable `AGENTS.md` instructions.
2. Inspect the complete diff for the requested target and enough surrounding code to understand
each changed path.
3. Identify concrete regressions introduced by the change. Continue through the whole diff after
finding the first issue.
4. Check the relevant tests and call sites to confirm that each finding is real and actionable.
For a base-branch review, compare the changes that would actually merge rather than diffing
directly against the branch tip. Resolve the comparison ref to the branch's upstream when that
upstream exists and is ahead of the local branch; otherwise use the local branch. Run
`git merge-base HEAD <comparison-ref>`, then inspect `git diff <merge-base-sha>`. If the local
branch cannot be resolved, try its configured upstream explicitly before reporting that the target
is unavailable.
Flag an issue only when all of these are true:
- It affects correctness, security, performance, or maintainability in a meaningful way.
- It is discrete and actionable.
- It was introduced by the reviewed change.
- The affected scenario or call path can be demonstrated from the code.
- The author would probably fix it if they knew about it.
Do not flag speculative concerns, pre-existing problems, intentional behavior changes, or style
nits that do not obscure the code.
## Write the result
Present findings first, ordered by severity. Use one entry per issue in this form:
`[P1] Imperative finding title — path/to/file.rs:line`
Follow the title with one short paragraph explaining the affected scenario and why the behavior is
wrong. Keep the cited range as small as possible and make sure it overlaps the reviewed diff.
Use these priorities:
- `P0`: universal release blocker or critical failure.
- `P1`: urgent defect that should be fixed next.
- `P2`: ordinary defect that should be fixed.
- `P3`: low-impact issue that is still worth fixing.
If there are no qualifying findings, say `No findings.` Do not invent a finding to fill the result.
After the findings, add a brief overall assessment and mention any material test gaps or residual
risks.

6
.codex-home/skills/.system/review-agent/agents/openai.yaml Normal file
View file

@ -0,0 +1,6 @@
interface:
display_name: "Review Agent"
short_description: "Find actionable bugs in code changes"
default_prompt: "Use $review-agent to review the requested code changes and return actionable findings."
policy:
allow_implicit_invocation: false

416
.codex-home/skills/.system/skill-creator/SKILL.md Normal file
View file

@ -0,0 +1,416 @@
---
name: skill-creator
description: Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
metadata:
short-description: Create or update a skill
---
# Skill Creator
This skill provides guidance for creating effective skills.
## About Skills
Skills are modular, self-contained folders that extend Codex's capabilities by providing
specialized knowledge, workflows, and tools. Think of them as "onboarding guides" for specific
domains or tasks—they transform Codex from a general-purpose agent into a specialized agent
equipped with procedural knowledge that no model can fully possess.
### What Skills Provide
1. Specialized workflows - Multi-step procedures for specific domains
2. Tool integrations - Instructions for working with specific file formats or APIs
3. Domain expertise - Company-specific knowledge, schemas, business logic
4. Bundled resources - Scripts, references, and assets for complex and repetitive tasks
## Core Principles
### Concise is Key
The context window is a public good. Skills share the context window with everything else Codex needs: system prompt, conversation history, other Skills' metadata, and the actual user request.
**Default assumption: Codex is already very smart.** Only add context Codex doesn't already have. Challenge each piece of information: "Does Codex really need this explanation?" and "Does this paragraph justify its token cost?"
Prefer concise examples over verbose explanations.
### Set Appropriate Degrees of Freedom
Match the level of specificity to the task's fragility and variability:
**High freedom (text-based instructions)**: Use when multiple approaches are valid, decisions depend on context, or heuristics guide the approach.
**Medium freedom (pseudocode or scripts with parameters)**: Use when a preferred pattern exists, some variation is acceptable, or configuration affects behavior.
**Low freedom (specific scripts, few parameters)**: Use when operations are fragile and error-prone, consistency is critical, or a specific sequence must be followed.
Think of Codex as exploring a path: a narrow bridge with cliffs needs specific guardrails (low freedom), while an open field allows many routes (high freedom).
### Protect Validation Integrity
You may use subagents during iteration to validate whether a skill works on realistic tasks or whether a suspected problem is real. This is most useful when you want an independent pass on the skill's behavior, outputs, or failure modes after a revision. Only do this when it is possible to start new subagents.
When using subagents for validation, treat that as an evaluation surface. The goal is to learn whether the skill generalizes, not whether another agent can reconstruct the answer from leaked context.
Prefer raw artifacts such as example prompts, outputs, diffs, logs, or traces. Give the minimum task-local context needed to perform the validation. Avoid passing the intended answer, suspected bug, intended fix, or your prior conclusions unless the validation explicitly requires them.
### Anatomy of a Skill
Every skill consists of a required SKILL.md file and optional bundled resources:
```
skill-name/
├── SKILL.md (required)
│ ├── YAML frontmatter metadata (required)
│ │ ├── name: (required)
│ │ └── description: (required)
│ └── Markdown instructions (required)
├── agents/ (recommended)
│ └── openai.yaml - UI metadata for skill lists and chips
└── Bundled Resources (optional)
├── scripts/ - Executable code (Python/Bash/etc.)
├── references/ - Documentation intended to be loaded into context as needed
└── assets/ - Files used in output (templates, icons, fonts, etc.)
```
#### SKILL.md (required)
Every SKILL.md consists of:
- **Frontmatter** (YAML): Contains `name` and `description` fields. These are the only fields that Codex reads to determine when the skill gets used, thus it is very important to be clear and comprehensive in describing what the skill is, and when it should be used.
- **Body** (Markdown): Instructions and guidance for using the skill. Only loaded AFTER the skill triggers (if at all).
#### Agents metadata (recommended)
- UI-facing metadata for skill lists and chips
- Read references/openai_yaml.md before generating values and follow its descriptions and constraints
- Create: human-facing `display_name`, `short_description`, and `default_prompt` by reading the skill
- Generate deterministically by passing the values as `--interface key=value` to `scripts/generate_openai_yaml.py` or `scripts/init_skill.py`
- On updates: validate `agents/openai.yaml` still matches SKILL.md; regenerate if stale
- Only include other optional interface fields (icons, brand color) if explicitly provided
- See references/openai_yaml.md for field definitions and examples
#### Bundled Resources (optional)
##### Scripts (`scripts/`)
Executable code (Python/Bash/etc.) for tasks that require deterministic reliability or are repeatedly rewritten.
- **When to include**: When the same code is being rewritten repeatedly or deterministic reliability is needed
- **Example**: `scripts/rotate_pdf.py` for PDF rotation tasks
- **Benefits**: Token efficient, deterministic, may be executed without loading into context
- **Note**: Scripts may still need to be read by Codex for patching or environment-specific adjustments
##### References (`references/`)
Documentation and reference material intended to be loaded as needed into context to inform Codex's process and thinking.
- **When to include**: For documentation that Codex should reference while working
- **Examples**: `references/finance.md` for financial schemas, `references/mnda.md` for company NDA template, `references/policies.md` for company policies, `references/api_docs.md` for API specifications
- **Use cases**: Database schemas, API documentation, domain knowledge, company policies, detailed workflow guides
- **Benefits**: Keeps SKILL.md lean, loaded only when Codex determines it's needed
- **Best practice**: If files are large (>10k words), include grep search patterns in SKILL.md
- **Avoid duplication**: Information should live in either SKILL.md or references files, not both. Prefer references files for detailed information unless it's truly core to the skill—this keeps SKILL.md lean while making information discoverable without hogging the context window. Keep only essential procedural instructions and workflow guidance in SKILL.md; move detailed reference material, schemas, and examples to references files.
##### Assets (`assets/`)
Files not intended to be loaded into context, but rather used within the output Codex produces.
- **When to include**: When the skill needs files that will be used in the final output
- **Examples**: `assets/logo.png` for brand assets, `assets/slides.pptx` for PowerPoint templates, `assets/frontend-template/` for HTML/React boilerplate, `assets/font.ttf` for typography
- **Use cases**: Templates, images, icons, boilerplate code, fonts, sample documents that get copied or modified
- **Benefits**: Separates output resources from documentation, enables Codex to use files without loading them into context
#### What to Not Include in a Skill
A skill should only contain essential files that directly support its functionality. Do NOT create extraneous documentation or auxiliary files, including:
- README.md
- INSTALLATION_GUIDE.md
- QUICK_REFERENCE.md
- CHANGELOG.md
- etc.
The skill should only contain the information needed for an AI agent to do the job at hand. It should not contain auxiliary context about the process that went into creating it, setup and testing procedures, user-facing documentation, etc. Creating additional documentation files just adds clutter and confusion.
### Progressive Disclosure Design Principle
Skills use a three-level loading system to manage context efficiently:
1. **Metadata (name + description)** - Always in context (~100 words)
2. **SKILL.md body** - When skill triggers (<5k words)
3. **Bundled resources** - As needed by Codex (Unlimited because scripts can be executed without reading into context window)
#### Progressive Disclosure Patterns
Keep SKILL.md body to the essentials and under 500 lines to minimize context bloat. Split content into separate files when approaching this limit. When splitting out content into other files, it is very important to reference them from SKILL.md and describe clearly when to read them, to ensure the reader of the skill knows they exist and when to use them.
**Key principle:** When a skill supports multiple variations, frameworks, or options, keep only the core workflow and selection guidance in SKILL.md. Move variant-specific details (patterns, examples, configuration) into separate reference files.
**Pattern 1: High-level guide with references**
```markdown
# PDF Processing
## Quick start
Extract text with pdfplumber:
[code example]
## Advanced features
- **Form filling**: See [FORMS.md](FORMS.md) for complete guide
- **API reference**: See [REFERENCE.md](REFERENCE.md) for all methods
- **Examples**: See [EXAMPLES.md](EXAMPLES.md) for common patterns
```
Codex loads FORMS.md, REFERENCE.md, or EXAMPLES.md only when needed.
**Pattern 2: Domain-specific organization**
For Skills with multiple domains, organize content by domain to avoid loading irrelevant context:
```
bigquery-skill/
├── SKILL.md (overview and navigation)
└── reference/
├── finance.md (revenue, billing metrics)
├── sales.md (opportunities, pipeline)
├── product.md (API usage, features)
└── marketing.md (campaigns, attribution)
```
When a user asks about sales metrics, Codex only reads sales.md.
Similarly, for skills supporting multiple frameworks or variants, organize by variant:
```
cloud-deploy/
├── SKILL.md (workflow + provider selection)
└── references/
├── aws.md (AWS deployment patterns)
├── gcp.md (GCP deployment patterns)
└── azure.md (Azure deployment patterns)
```
When the user chooses AWS, Codex only reads aws.md.
**Pattern 3: Conditional details**
Show basic content, link to advanced content:
```markdown
# DOCX Processing
## Creating documents
Use docx-js for new documents. See [DOCX-JS.md](DOCX-JS.md).
## Editing documents
For simple edits, modify the XML directly.
**For tracked changes**: See [REDLINING.md](REDLINING.md)
**For OOXML details**: See [OOXML.md](OOXML.md)
```
Codex reads REDLINING.md or OOXML.md only when the user needs those features.
**Important guidelines:**
- **Avoid deeply nested references** - Keep references one level deep from SKILL.md. All reference files should link directly from SKILL.md.
- **Structure longer reference files** - For files longer than 100 lines, include a table of contents at the top so Codex can see the full scope when previewing.
## Skill Creation Process
Skill creation involves these steps:
1. Understand the skill with concrete examples
2. Plan reusable skill contents (scripts, references, assets)
3. Initialize the skill (run init_skill.py)
4. Edit the skill (implement resources and write SKILL.md)
5. Validate the skill (run quick_validate.py)
6. Iterate based on real usage and forward-test complex skills.
Follow these steps in order, skipping only if there is a clear reason why they are not applicable.
### Skill Naming
- Use lowercase letters, digits, and hyphens only; normalize user-provided titles to hyphen-case (e.g., "Plan Mode" -> `plan-mode`).
- When generating names, generate a name under 64 characters (letters, digits, hyphens).
- Prefer short, verb-led phrases that describe the action.
- Namespace by tool when it improves clarity or triggering (e.g., `gh-address-comments`, `linear-address-issue`).
- Name the skill folder exactly after the skill name.
### Step 1: Understanding the Skill with Concrete Examples
Skip this step only when the skill's usage patterns are already clearly understood. It remains valuable even when working with an existing skill.
To create an effective skill, clearly understand concrete examples of how the skill will be used. This understanding can come from either direct user examples or generated examples that are validated with user feedback.
For example, when building an image-editor skill, relevant questions include:
- "What functionality should the image-editor skill support? Editing, rotating, anything else?"
- "Can you give some examples of how this skill would be used?"
- "I can imagine users asking for things like 'Remove the red-eye from this image' or 'Rotate this image'. Are there other ways you imagine this skill being used?"
- "What would a user say that should trigger this skill?"
- "Where should I create this skill? If you do not have a preference, I will place it in `$CODEX_HOME/skills` (or `~/.codex/skills` when `CODEX_HOME` is unset) so Codex can discover it automatically."
To avoid overwhelming users, avoid asking too many questions in a single message. Start with the most important questions and follow up as needed for better effectiveness.
Conclude this step when there is a clear sense of the functionality the skill should support.
### Step 2: Planning the Reusable Skill Contents
To turn concrete examples into an effective skill, analyze each example by:
1. Considering how to execute on the example from scratch
2. Identifying what scripts, references, and assets would be helpful when executing these workflows repeatedly
Example: When building a `pdf-editor` skill to handle queries like "Help me rotate this PDF," the analysis shows:
1. Rotating a PDF requires re-writing the same code each time
2. A `scripts/rotate_pdf.py` script would be helpful to store in the skill
Example: When designing a `frontend-webapp-builder` skill for queries like "Build me a todo app" or "Build me a dashboard to track my steps," the analysis shows:
1. Writing a frontend webapp requires the same boilerplate HTML/React each time
2. An `assets/hello-world/` template containing the boilerplate HTML/React project files would be helpful to store in the skill
Example: When building a `big-query` skill to handle queries like "How many users have logged in today?" the analysis shows:
1. Querying BigQuery requires re-discovering the table schemas and relationships each time
2. A `references/schema.md` file documenting the table schemas would be helpful to store in the skill
To establish the skill's contents, analyze each concrete example to create a list of the reusable resources to include: scripts, references, and assets.
### Step 3: Initializing the Skill
At this point, it is time to actually create the skill.
Skip this step only if the skill being developed already exists. In this case, continue to the next step.
Before running `init_skill.py`, ask where the user wants the skill created. If they do not specify a location, default to `$CODEX_HOME/skills`; when `CODEX_HOME` is unset, fall back to `~/.codex/skills` so the skill is auto-discovered.
When creating a new skill from scratch, always run the `init_skill.py` script. The script conveniently generates a new template skill directory that automatically includes everything a skill requires, making the skill creation process much more efficient and reliable.
Usage:
```bash
scripts/init_skill.py <skill-name> --path <output-directory> [--resources scripts,references,assets] [--examples]
```
Examples:
```bash
scripts/init_skill.py my-skill --path "${CODEX_HOME:-$HOME/.codex}/skills"
scripts/init_skill.py my-skill --path "${CODEX_HOME:-$HOME/.codex}/skills" --resources scripts,references
scripts/init_skill.py my-skill --path ~/work/skills --resources scripts --examples
```
The script:
- Creates the skill directory at the specified path
- Generates a SKILL.md template with proper frontmatter and TODO placeholders
- Creates `agents/openai.yaml` using agent-generated `display_name`, `short_description`, and `default_prompt` passed via `--interface key=value`
- Optionally creates resource directories based on `--resources`
- Optionally adds example files when `--examples` is set
After initialization, customize the SKILL.md and add resources as needed. If you used `--examples`, replace or delete placeholder files.
Generate `display_name`, `short_description`, and `default_prompt` by reading the skill, then pass them as `--interface key=value` to `init_skill.py` or regenerate with:
```bash
scripts/generate_openai_yaml.py <path/to/skill-folder> --interface key=value
```
Only include other optional interface fields when the user explicitly provides them. For full field descriptions and examples, see references/openai_yaml.md.
### Step 4: Edit the Skill
When editing the (newly-generated or existing) skill, remember that the skill is being created for another instance of Codex to use. Include information that would be beneficial and non-obvious to Codex. Consider what procedural knowledge, domain-specific details, or reusable assets would help another Codex instance execute these tasks more effectively.
After substantial revisions, or if the skill is particularly tricky, you should use subagents to forward-test the skill on realistic tasks or artifacts. When doing so, pass the artifact under validation rather than your diagnosis of what is wrong, and keep the prompt generic enough that success depends on transferable reasoning rather than hidden ground truth.
#### Start with Reusable Skill Contents
To begin implementation, start with the reusable resources identified above: `scripts/`, `references/`, and `assets/` files. Note that this step may require user input. For example, when implementing a `brand-guidelines` skill, the user may need to provide brand assets or templates to store in `assets/`, or documentation to store in `references/`.
Added scripts must be tested by actually running them to ensure there are no bugs and that the output matches what is expected. If there are many similar scripts, only a representative sample needs to be tested to ensure confidence that they all work while balancing time to completion.
If you used `--examples`, delete any placeholder files that are not needed for the skill. Only create resource directories that are actually required.
#### Update SKILL.md
**Writing Guidelines:** Always use imperative/infinitive form.
##### Frontmatter
Write the YAML frontmatter with `name` and `description`:
- `name`: The skill name
- `description`: This is the primary triggering mechanism for your skill, and helps Codex understand when to use the skill.
- Include both what the Skill does and specific triggers/contexts for when to use it.
- Include all "when to use" information here - Not in the body. The body is only loaded after triggering, so "When to Use This Skill" sections in the body are not helpful to Codex.
- Example description for a `docx` skill: "Comprehensive document creation, editing, and analysis with support for tracked changes, comments, formatting preservation, and text extraction. Use when Codex needs to work with professional documents (.docx files) for: (1) Creating new documents, (2) Modifying or editing content, (3) Working with tracked changes, (4) Adding comments, or any other document tasks"
Do not include any other fields in YAML frontmatter.
##### Body
Write instructions for using the skill and its bundled resources.
### Step 5: Validate the Skill
Once development of the skill is complete, validate the skill folder to catch basic issues early:
```bash
scripts/quick_validate.py <path/to/skill-folder>
```
The validation script checks YAML frontmatter format, required fields, and naming rules. If validation fails, fix the reported issues and run the command again.
### Step 6: Iterate
After testing the skill, you may detect the skill is complex enough that it requires forward-testing; or users may request improvements.
User testing often this happens right after using the skill, with fresh context of how the skill performed.
**Forward-testing and iteration workflow:**
1. Use the skill on real tasks
2. Notice struggles or inefficiencies
3. Identify how SKILL.md or bundled resources should be updated
4. Implement changes and test again
5. Forward-test if it is reasonable and appropriate
## Forward-testing
To forward-test, launch subagents as a way to stress test the skill with minimal context.
Subagents should *not* know that they are being asked to test the skill. They should be treated as
an agent asked to perform a task by the user. Prompts to subagents should look like:
`Use $skill-x at /path/to/skill-x to solve problem y`
Not:
`Review the skill at /path/to/skill-x; pretend a user asks you to...`
Decision rule for forward-testing:
- Err on the side of forward-testing
- Ask for approval if you think there's a risk that forward-testing would:
* take a long time,
* require additional approvals from the user, or
* modify live production systems
In these cases, show the user your proposed prompt and request (1) a yes/no decision, and
(2) any suggested modifictions.
Considerations when forward-testing:
- use fresh threads for independent passes
- pass the skill, and a request in a similar way the user would.
- pass raw artifacts, not your conclusions
- avoid showing expected answers or intended fixes
- rebuild context from source artifacts after each iteration
- review the subagent's output and reasoning and emitted artifacts
- avoid leaving artifacts the agent can find on disk between iterations;
clean up subagents' artifacts to avoid additional contamination.
If forward-testing only succeeds when subagents see leaked context, tighten the skill or the
forward-testing setup before trusting the result.

5
.codex-home/skills/.system/skill-creator/agents/openai.yaml Normal file
View file

@ -0,0 +1,5 @@
interface:
display_name: "Skill Creator"
short_description: "Create or update a skill"
icon_small: "./assets/skill-creator-small.svg"
icon_large: "./assets/skill-creator.png"

3
.codex-home/skills/.system/skill-creator/assets/skill-creator-small.svg Normal file
View file

@ -0,0 +1,3 @@
<svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" fill="currentColor" viewBox="0 0 20 20">
<path fill="#0D0D0D" d="M12.03 4.113a3.612 3.612 0 0 1 5.108 5.108l-6.292 6.29c-.324.324-.56.561-.791.752l-.235.176c-.205.14-.422.261-.65.36l-.229.093a4.136 4.136 0 0 1-.586.16l-.764.134-2.394.4c-.142.024-.294.05-.423.06-.098.007-.232.01-.378-.026l-.149-.05a1.081 1.081 0 0 1-.521-.474l-.046-.093a1.104 1.104 0 0 1-.075-.527c.01-.129.035-.28.06-.422l.398-2.394c.1-.602.162-.987.295-1.35l.093-.23c.1-.228.22-.445.36-.65l.176-.235c.19-.232.428-.467.751-.79l6.292-6.292Zm-5.35 7.232c-.35.35-.534.535-.66.688l-.11.147a2.67 2.67 0 0 0-.24.433l-.062.154c-.08.22-.124.462-.232 1.112l-.398 2.394-.001.001h.003l2.393-.399.717-.126a2.63 2.63 0 0 0 .394-.105l.154-.063a2.65 2.65 0 0 0 .433-.24l.147-.11c.153-.126.339-.31.688-.66l4.988-4.988-3.227-3.226-4.987 4.988Zm9.517-6.291a2.281 2.281 0 0 0-3.225 0l-.364.362 3.226 3.227.363-.364c.89-.89.89-2.334 0-3.225ZM4.583 1.783a.3.3 0 0 1 .294.241c.117.585.347 1.092.707 1.48.357.385.859.668 1.549.783a.3.3 0 0 1 0 .592c-.69.115-1.192.398-1.549.783-.315.34-.53.77-.657 1.265l-.05.215a.3.3 0 0 1-.588 0c-.117-.585-.347-1.092-.707-1.48-.357-.384-.859-.668-1.549-.783a.3.3 0 0 1 0-.592c.69-.115 1.192-.398 1.549-.783.36-.388.59-.895.707-1.48l.015-.05a.3.3 0 0 1 .279-.19Z"/>
</svg>
Side by side

After

Width:  |  Height:  |  Size: 1.3 KiB

BIN
.codex-home/skills/.system/skill-creator/assets/skill-creator.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 1.5 KiB

202
.codex-home/skills/.system/skill-creator/license.txt Normal file
View file

@ -0,0 +1,202 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

49
.codex-home/skills/.system/skill-creator/references/openai_yaml.md Normal file
View file

@ -0,0 +1,49 @@
# openai.yaml fields (full example + descriptions)
`agents/openai.yaml` is an extended, product-specific config intended for the machine/harness to read, not the agent. Other product-specific config can also live in the `agents/` folder.
## Full example
```yaml
interface:
display_name: "Optional user-facing name"
short_description: "Optional user-facing description"
icon_small: "./assets/small-400px.png"
icon_large: "./assets/large-logo.svg"
brand_color: "#3B82F6"
default_prompt: "Optional surrounding prompt to use the skill with"
dependencies:
tools:
- type: "mcp"
value: "github"
description: "GitHub MCP server"
transport: "streamable_http"
url: "https://api.githubcopilot.com/mcp/"
policy:
allow_implicit_invocation: true
```
## Field descriptions and constraints
Top-level constraints:
- Quote all string values.
- Keep keys unquoted.
- For `interface.default_prompt`: generate a helpful, short (typically 1 sentence) example starting prompt based on the skill. It must explicitly mention the skill as `$skill-name` (e.g., "Use $skill-name-here to draft a concise weekly status update.").
- `interface.display_name`: Human-facing title shown in UI skill lists and chips.
- `interface.short_description`: Human-facing short UI blurb (25–64 chars) for quick scanning.
- `interface.icon_small`: Path to a small icon asset (relative to skill dir). Default to `./assets/` and place icons in the skill's `assets/` folder.
- `interface.icon_large`: Path to a larger logo asset (relative to skill dir). Default to `./assets/` and place icons in the skill's `assets/` folder.
- `interface.brand_color`: Hex color used for UI accents (e.g., badges).
- `interface.default_prompt`: Default prompt snippet inserted when invoking the skill.
- `dependencies.tools[].type`: Dependency category. Only `mcp` is supported for now.
- `dependencies.tools[].value`: Identifier of the tool or dependency.
- `dependencies.tools[].description`: Human-readable explanation of the dependency.
- `dependencies.tools[].transport`: Connection type when `type` is `mcp`.
- `dependencies.tools[].url`: MCP server URL when `type` is `mcp`.
- `policy.allow_implicit_invocation`: When false, the skill is not injected into
the model context by default, but can still be invoked explicitly via `$skill`.
Defaults to true.

226
.codex-home/skills/.system/skill-creator/scripts/generate_openai_yaml.py Normal file
View file

@ -0,0 +1,226 @@
#!/usr/bin/env python3
"""
OpenAI YAML Generator - Creates agents/openai.yaml for a skill folder.
Usage:
generate_openai_yaml.py <skill_dir> [--name <skill_name>] [--interface key=value]
"""
import argparse
import re
import sys
from pathlib import Path
ACRONYMS = {
"GH",
"MCP",
"API",
"CI",
"CLI",
"LLM",
"PDF",
"PR",
"UI",
"URL",
"SQL",
}
BRANDS = {
"openai": "OpenAI",
"openapi": "OpenAPI",
"github": "GitHub",
"pagerduty": "PagerDuty",
"datadog": "DataDog",
"sqlite": "SQLite",
"fastapi": "FastAPI",
}
SMALL_WORDS = {"and", "or", "to", "up", "with"}
ALLOWED_INTERFACE_KEYS = {
"display_name",
"short_description",
"icon_small",
"icon_large",
"brand_color",
"default_prompt",
}
def yaml_quote(value):
escaped = value.replace("\\", "\\\\").replace('"', '\\"').replace("\n", "\\n")
return f'"{escaped}"'
def format_display_name(skill_name):
words = [word for word in skill_name.split("-") if word]
formatted = []
for index, word in enumerate(words):
lower = word.lower()
upper = word.upper()
if upper in ACRONYMS:
formatted.append(upper)
continue
if lower in BRANDS:
formatted.append(BRANDS[lower])
continue
if index > 0 and lower in SMALL_WORDS:
formatted.append(lower)
continue
formatted.append(word.capitalize())
return " ".join(formatted)
def generate_short_description(display_name):
description = f"Help with {display_name} tasks"
if len(description) < 25:
description = f"Help with {display_name} tasks and workflows"
if len(description) < 25:
description = f"Help with {display_name} tasks with guidance"
if len(description) > 64:
description = f"Help with {display_name}"
if len(description) > 64:
description = f"{display_name} helper"
if len(description) > 64:
description = f"{display_name} tools"
if len(description) > 64:
suffix = " helper"
max_name_length = 64 - len(suffix)
trimmed = display_name[:max_name_length].rstrip()
description = f"{trimmed}{suffix}"
if len(description) > 64:
description = description[:64].rstrip()
if len(description) < 25:
description = f"{description} workflows"
if len(description) > 64:
description = description[:64].rstrip()
return description
def read_frontmatter_name(skill_dir):
skill_md = Path(skill_dir) / "SKILL.md"
if not skill_md.exists():
print(f"[ERROR] SKILL.md not found in {skill_dir}")
return None
content = skill_md.read_text()
match = re.match(r"^---\n(.*?)\n---", content, re.DOTALL)
if not match:
print("[ERROR] Invalid SKILL.md frontmatter format.")
return None
frontmatter_text = match.group(1)
import yaml
try:
frontmatter = yaml.safe_load(frontmatter_text)
except yaml.YAMLError as exc:
print(f"[ERROR] Invalid YAML frontmatter: {exc}")
return None
if not isinstance(frontmatter, dict):
print("[ERROR] Frontmatter must be a YAML dictionary.")
return None
name = frontmatter.get("name", "")
if not isinstance(name, str) or not name.strip():
print("[ERROR] Frontmatter 'name' is missing or invalid.")
return None
return name.strip()
def parse_interface_overrides(raw_overrides):
overrides = {}
optional_order = []
for item in raw_overrides:
if "=" not in item:
print(f"[ERROR] Invalid interface override '{item}'. Use key=value.")
return None, None
key, value = item.split("=", 1)
key = key.strip()
value = value.strip()
if not key:
print(f"[ERROR] Invalid interface override '{item}'. Key is empty.")
return None, None
if key not in ALLOWED_INTERFACE_KEYS:
allowed = ", ".join(sorted(ALLOWED_INTERFACE_KEYS))
print(f"[ERROR] Unknown interface field '{key}'. Allowed: {allowed}")
return None, None
overrides[key] = value
if key not in ("display_name", "short_description") and key not in optional_order:
optional_order.append(key)
return overrides, optional_order
def write_openai_yaml(skill_dir, skill_name, raw_overrides):
overrides, optional_order = parse_interface_overrides(raw_overrides)
if overrides is None:
return None
display_name = overrides.get("display_name") or format_display_name(skill_name)
short_description = overrides.get("short_description") or generate_short_description(display_name)
if not (25 <= len(short_description) <= 64):
print(
"[ERROR] short_description must be 25-64 characters "
f"(got {len(short_description)})."
)
return None
interface_lines = [
"interface:",
f" display_name: {yaml_quote(display_name)}",
f" short_description: {yaml_quote(short_description)}",
]
for key in optional_order:
value = overrides.get(key)
if value is not None:
interface_lines.append(f" {key}: {yaml_quote(value)}")
agents_dir = Path(skill_dir) / "agents"
agents_dir.mkdir(parents=True, exist_ok=True)
output_path = agents_dir / "openai.yaml"
output_path.write_text("\n".join(interface_lines) + "\n")
print(f"[OK] Created agents/openai.yaml")
return output_path
def main():
parser = argparse.ArgumentParser(
description="Create agents/openai.yaml for a skill directory.",
)
parser.add_argument("skill_dir", help="Path to the skill directory")
parser.add_argument(
"--name",
help="Skill name override (defaults to SKILL.md frontmatter)",
)
parser.add_argument(
"--interface",
action="append",
default=[],
help="Interface override in key=value format (repeatable)",
)
args = parser.parse_args()
skill_dir = Path(args.skill_dir).resolve()
if not skill_dir.exists():
print(f"[ERROR] Skill directory not found: {skill_dir}")
sys.exit(1)
if not skill_dir.is_dir():
print(f"[ERROR] Path is not a directory: {skill_dir}")
sys.exit(1)
skill_name = args.name or read_frontmatter_name(skill_dir)
if not skill_name:
sys.exit(1)
result = write_openai_yaml(skill_dir, skill_name, args.interface)
if result:
sys.exit(0)
sys.exit(1)
if __name__ == "__main__":
main()

400
.codex-home/skills/.system/skill-creator/scripts/init_skill.py Normal file
View file

@ -0,0 +1,400 @@
#!/usr/bin/env python3
"""
Skill Initializer - Creates a new skill from template
Usage:
init_skill.py <skill-name> --path <path> [--resources scripts,references,assets] [--examples] [--interface key=value]
Examples:
init_skill.py my-new-skill --path skills/public
init_skill.py my-new-skill --path skills/public --resources scripts,references
init_skill.py my-api-helper --path skills/private --resources scripts --examples
init_skill.py custom-skill --path /custom/location
init_skill.py my-skill --path skills/public --interface short_description="Short UI label"
"""
import argparse
import re
import sys
from pathlib import Path
from generate_openai_yaml import write_openai_yaml
MAX_SKILL_NAME_LENGTH = 64
ALLOWED_RESOURCES = {"scripts", "references", "assets"}
SKILL_TEMPLATE = """---
name: {skill_name}
description: [TODO: Complete and informative explanation of what the skill does and when to use it. Include WHEN to use this skill - specific scenarios, file types, or tasks that trigger it.]
---
# {skill_title}
## Overview
[TODO: 1-2 sentences explaining what this skill enables]
## Structuring This Skill
[TODO: Choose the structure that best fits this skill's purpose. Common patterns:
**1. Workflow-Based** (best for sequential processes)
- Works well when there are clear step-by-step procedures
- Example: DOCX skill with "Workflow Decision Tree" -> "Reading" -> "Creating" -> "Editing"
- Structure: ## Overview -> ## Workflow Decision Tree -> ## Step 1 -> ## Step 2...
**2. Task-Based** (best for tool collections)
- Works well when the skill offers different operations/capabilities
- Example: PDF skill with "Quick Start" -> "Merge PDFs" -> "Split PDFs" -> "Extract Text"
- Structure: ## Overview -> ## Quick Start -> ## Task Category 1 -> ## Task Category 2...
**3. Reference/Guidelines** (best for standards or specifications)
- Works well for brand guidelines, coding standards, or requirements
- Example: Brand styling with "Brand Guidelines" -> "Colors" -> "Typography" -> "Features"
- Structure: ## Overview -> ## Guidelines -> ## Specifications -> ## Usage...
**4. Capabilities-Based** (best for integrated systems)
- Works well when the skill provides multiple interrelated features
- Example: Product Management with "Core Capabilities" -> numbered capability list
- Structure: ## Overview -> ## Core Capabilities -> ### 1. Feature -> ### 2. Feature...
Patterns can be mixed and matched as needed. Most skills combine patterns (e.g., start with task-based, add workflow for complex operations).
Delete this entire "Structuring This Skill" section when done - it's just guidance.]
## [TODO: Replace with the first main section based on chosen structure]
[TODO: Add content here. See examples in existing skills:
- Code samples for technical skills
- Decision trees for complex workflows
- Concrete examples with realistic user requests
- References to scripts/templates/references as needed]
## Resources (optional)
Create only the resource directories this skill actually needs. Delete this section if no resources are required.
### scripts/
Executable code (Python/Bash/etc.) that can be run directly to perform specific operations.
**Examples from other skills:**
- PDF skill: `fill_fillable_fields.py`, `extract_form_field_info.py` - utilities for PDF manipulation
- DOCX skill: `document.py`, `utilities.py` - Python modules for document processing
**Appropriate for:** Python scripts, shell scripts, or any executable code that performs automation, data processing, or specific operations.
**Note:** Scripts may be executed without loading into context, but can still be read by Codex for patching or environment adjustments.
### references/
Documentation and reference material intended to be loaded into context to inform Codex's process and thinking.
**Examples from other skills:**
- Product management: `communication.md`, `context_building.md` - detailed workflow guides
- BigQuery: API reference documentation and query examples
- Finance: Schema documentation, company policies
**Appropriate for:** In-depth documentation, API references, database schemas, comprehensive guides, or any detailed information that Codex should reference while working.
### assets/
Files not intended to be loaded into context, but rather used within the output Codex produces.
**Examples from other skills:**
- Brand styling: PowerPoint template files (.pptx), logo files
- Frontend builder: HTML/React boilerplate project directories
- Typography: Font files (.ttf, .woff2)
**Appropriate for:** Templates, boilerplate code, document templates, images, icons, fonts, or any files meant to be copied or used in the final output.
---
**Not every skill requires all three types of resources.**
"""
EXAMPLE_SCRIPT = '''#!/usr/bin/env python3
"""
Example helper script for {skill_name}
This is a placeholder script that can be executed directly.
Replace with actual implementation or delete if not needed.
Example real scripts from other skills:
- pdf/scripts/fill_fillable_fields.py - Fills PDF form fields
- pdf/scripts/convert_pdf_to_images.py - Converts PDF pages to images
"""
def main():
print("This is an example script for {skill_name}")
# TODO: Add actual script logic here
# This could be data processing, file conversion, API calls, etc.
if __name__ == "__main__":
main()
'''
EXAMPLE_REFERENCE = """# Reference Documentation for {skill_title}
This is a placeholder for detailed reference documentation.
Replace with actual reference content or delete if not needed.
Example real reference docs from other skills:
- product-management/references/communication.md - Comprehensive guide for status updates
- product-management/references/context_building.md - Deep-dive on gathering context
- bigquery/references/ - API references and query examples
## When Reference Docs Are Useful
Reference docs are ideal for:
- Comprehensive API documentation
- Detailed workflow guides
- Complex multi-step processes
- Information too lengthy for main SKILL.md
- Content that's only needed for specific use cases
## Structure Suggestions
### API Reference Example
- Overview
- Authentication
- Endpoints with examples
- Error codes
- Rate limits
### Workflow Guide Example
- Prerequisites
- Step-by-step instructions
- Common patterns
- Troubleshooting
- Best practices
"""
EXAMPLE_ASSET = """# Example Asset File
This placeholder represents where asset files would be stored.
Replace with actual asset files (templates, images, fonts, etc.) or delete if not needed.
Asset files are NOT intended to be loaded into context, but rather used within
the output Codex produces.
Example asset files from other skills:
- Brand guidelines: logo.png, slides_template.pptx
- Frontend builder: hello-world/ directory with HTML/React boilerplate
- Typography: custom-font.ttf, font-family.woff2
- Data: sample_data.csv, test_dataset.json
## Common Asset Types
- Templates: .pptx, .docx, boilerplate directories
- Images: .png, .jpg, .svg, .gif
- Fonts: .ttf, .otf, .woff, .woff2
- Boilerplate code: Project directories, starter files
- Icons: .ico, .svg
- Data files: .csv, .json, .xml, .yaml
Note: This is a text placeholder. Actual assets can be any file type.
"""
def normalize_skill_name(skill_name):
"""Normalize a skill name to lowercase hyphen-case."""
normalized = skill_name.strip().lower()
normalized = re.sub(r"[^a-z0-9]+", "-", normalized)
normalized = normalized.strip("-")
normalized = re.sub(r"-{2,}", "-", normalized)
return normalized
def title_case_skill_name(skill_name):
"""Convert hyphenated skill name to Title Case for display."""
return " ".join(word.capitalize() for word in skill_name.split("-"))
def parse_resources(raw_resources):
if not raw_resources:
return []
resources = [item.strip() for item in raw_resources.split(",") if item.strip()]
invalid = sorted({item for item in resources if item not in ALLOWED_RESOURCES})
if invalid:
allowed = ", ".join(sorted(ALLOWED_RESOURCES))
print(f"[ERROR] Unknown resource type(s): {', '.join(invalid)}")
print(f" Allowed: {allowed}")
sys.exit(1)
deduped = []
seen = set()
for resource in resources:
if resource not in seen:
deduped.append(resource)
seen.add(resource)
return deduped
def create_resource_dirs(skill_dir, skill_name, skill_title, resources, include_examples):
for resource in resources:
resource_dir = skill_dir / resource
resource_dir.mkdir(exist_ok=True)
if resource == "scripts":
if include_examples:
example_script = resource_dir / "example.py"
example_script.write_text(EXAMPLE_SCRIPT.format(skill_name=skill_name))
example_script.chmod(0o755)
print("[OK] Created scripts/example.py")
else:
print("[OK] Created scripts/")
elif resource == "references":
if include_examples:
example_reference = resource_dir / "api_reference.md"
example_reference.write_text(EXAMPLE_REFERENCE.format(skill_title=skill_title))
print("[OK] Created references/api_reference.md")
else:
print("[OK] Created references/")
elif resource == "assets":
if include_examples:
example_asset = resource_dir / "example_asset.txt"
example_asset.write_text(EXAMPLE_ASSET)
print("[OK] Created assets/example_asset.txt")
else:
print("[OK] Created assets/")
def init_skill(skill_name, path, resources, include_examples, interface_overrides):
"""
Initialize a new skill directory with template SKILL.md.
Args:
skill_name: Name of the skill
path: Path where the skill directory should be created
resources: Resource directories to create
include_examples: Whether to create example files in resource directories
Returns:
Path to created skill directory, or None if error
"""
# Determine skill directory path
skill_dir = Path(path).resolve() / skill_name
# Check if directory already exists
if skill_dir.exists():
print(f"[ERROR] Skill directory already exists: {skill_dir}")
return None
# Create skill directory
try:
skill_dir.mkdir(parents=True, exist_ok=False)
print(f"[OK] Created skill directory: {skill_dir}")
except Exception as e:
print(f"[ERROR] Error creating directory: {e}")
return None
# Create SKILL.md from template
skill_title = title_case_skill_name(skill_name)
skill_content = SKILL_TEMPLATE.format(skill_name=skill_name, skill_title=skill_title)
skill_md_path = skill_dir / "SKILL.md"
try:
skill_md_path.write_text(skill_content)
print("[OK] Created SKILL.md")
except Exception as e:
print(f"[ERROR] Error creating SKILL.md: {e}")
return None
# Create agents/openai.yaml
try:
result = write_openai_yaml(skill_dir, skill_name, interface_overrides)
if not result:
return None
except Exception as e:
print(f"[ERROR] Error creating agents/openai.yaml: {e}")
return None
# Create resource directories if requested
if resources:
try:
create_resource_dirs(skill_dir, skill_name, skill_title, resources, include_examples)
except Exception as e:
print(f"[ERROR] Error creating resource directories: {e}")
return None
# Print next steps
print(f"\n[OK] Skill '{skill_name}' initialized successfully at {skill_dir}")
print("\nNext steps:")
print("1. Edit SKILL.md to complete the TODO items and update the description")
if resources:
if include_examples:
print("2. Customize or delete the example files in scripts/, references/, and assets/")
else:
print("2. Add resources to scripts/, references/, and assets/ as needed")
else:
print("2. Create resource directories only if needed (scripts/, references/, assets/)")
print("3. Update agents/openai.yaml if the UI metadata should differ")
print("4. Run the validator when ready to check the skill structure")
print(
"5. Forward-test complex skills with realistic user requests to ensure they work as intended"
)
return skill_dir
def main():
parser = argparse.ArgumentParser(
description="Create a new skill directory with a SKILL.md template.",
)
parser.add_argument("skill_name", help="Skill name (normalized to hyphen-case)")
parser.add_argument("--path", required=True, help="Output directory for the skill")
parser.add_argument(
"--resources",
default="",
help="Comma-separated list: scripts,references,assets",
)
parser.add_argument(
"--examples",
action="store_true",
help="Create example files inside the selected resource directories",
)
parser.add_argument(
"--interface",
action="append",
default=[],
help="Interface override in key=value format (repeatable)",
)
args = parser.parse_args()
raw_skill_name = args.skill_name
skill_name = normalize_skill_name(raw_skill_name)
if not skill_name:
print("[ERROR] Skill name must include at least one letter or digit.")
sys.exit(1)
if len(skill_name) > MAX_SKILL_NAME_LENGTH:
print(
f"[ERROR] Skill name '{skill_name}' is too long ({len(skill_name)} characters). "
f"Maximum is {MAX_SKILL_NAME_LENGTH} characters."
)
sys.exit(1)
if skill_name != raw_skill_name:
print(f"Note: Normalized skill name from '{raw_skill_name}' to '{skill_name}'.")
resources = parse_resources(args.resources)
if args.examples and not resources:
print("[ERROR] --examples requires --resources to be set.")
sys.exit(1)
path = args.path
print(f"Initializing skill: {skill_name}")
print(f" Location: {path}")
if resources:
print(f" Resources: {', '.join(resources)}")
if args.examples:
print(" Examples: enabled")
else:
print(" Resources: none (create as needed)")
print()
result = init_skill(skill_name, path, resources, args.examples, args.interface)
if result:
sys.exit(0)
else:
sys.exit(1)
if __name__ == "__main__":
main()

101
.codex-home/skills/.system/skill-creator/scripts/quick_validate.py Normal file
View file

@ -0,0 +1,101 @@
#!/usr/bin/env python3
"""
Quick validation script for skills - minimal version
"""
import re
import sys
from pathlib import Path
import yaml
MAX_SKILL_NAME_LENGTH = 64
def validate_skill(skill_path):
"""Basic validation of a skill"""
skill_path = Path(skill_path)
skill_md = skill_path / "SKILL.md"
if not skill_md.exists():
return False, "SKILL.md not found"
content = skill_md.read_text()
if not content.startswith("---"):
return False, "No YAML frontmatter found"
match = re.match(r"^---\n(.*?)\n---", content, re.DOTALL)
if not match:
return False, "Invalid frontmatter format"
frontmatter_text = match.group(1)
try:
frontmatter = yaml.safe_load(frontmatter_text)
if not isinstance(frontmatter, dict):
return False, "Frontmatter must be a YAML dictionary"
except yaml.YAMLError as e:
return False, f"Invalid YAML in frontmatter: {e}"
allowed_properties = {"name", "description", "license", "allowed-tools", "metadata"}
unexpected_keys = set(frontmatter.keys()) - allowed_properties
if unexpected_keys:
allowed = ", ".join(sorted(allowed_properties))
unexpected = ", ".join(sorted(unexpected_keys))
return (
False,
f"Unexpected key(s) in SKILL.md frontmatter: {unexpected}. Allowed properties are: {allowed}",
)
if "name" not in frontmatter:
return False, "Missing 'name' in frontmatter"
if "description" not in frontmatter:
return False, "Missing 'description' in frontmatter"
name = frontmatter.get("name", "")
if not isinstance(name, str):
return False, f"Name must be a string, got {type(name).__name__}"
name = name.strip()
if name:
if not re.match(r"^[a-z0-9-]+$", name):
return (
False,
f"Name '{name}' should be hyphen-case (lowercase letters, digits, and hyphens only)",
)
if name.startswith("-") or name.endswith("-") or "--" in name:
return (
False,
f"Name '{name}' cannot start/end with hyphen or contain consecutive hyphens",
)
if len(name) > MAX_SKILL_NAME_LENGTH:
return (
False,
f"Name is too long ({len(name)} characters). "
f"Maximum is {MAX_SKILL_NAME_LENGTH} characters.",
)
description = frontmatter.get("description", "")
if not isinstance(description, str):
return False, f"Description must be a string, got {type(description).__name__}"
description = description.strip()
if description:
if "<" in description or ">" in description:
return False, "Description cannot contain angle brackets (< or >)"
if len(description) > 1024:
return (
False,
f"Description is too long ({len(description)} characters). Maximum is 1024 characters.",
)
return True, "Skill is valid!"
if __name__ == "__main__":
if len(sys.argv) != 2:
print("Usage: python quick_validate.py <skill_directory>")
sys.exit(1)
valid, message = validate_skill(sys.argv[1])
print(message)
sys.exit(0 if valid else 1)

202
.codex-home/skills/.system/skill-installer/LICENSE.txt Normal file
View file

@ -0,0 +1,202 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

58
.codex-home/skills/.system/skill-installer/SKILL.md Normal file
View file

@ -0,0 +1,58 @@
---
name: skill-installer
description: Install Codex skills into $CODEX_HOME/skills from a curated list or a GitHub repo path. Use when a user asks to list installable skills, install a curated skill, or install a skill from another repo (including private repos).
metadata:
short-description: Install curated skills from openai/skills or other repos
---
# Skill Installer
Helps install skills. By default these are from https://github.com/openai/skills/tree/main/skills/.curated, but users can also provide other locations. Experimental skills live in https://github.com/openai/skills/tree/main/skills/.experimental and can be installed the same way.
Use the helper scripts based on the task:
- List skills when the user asks what is available, or if the user uses this skill without specifying what to do. Default listing is `.curated`, but you can pass `--path skills/.experimental` when they ask about experimental skills.
- Install from the curated list when the user provides a skill name.
- Install from another repo when the user provides a GitHub repo/path (including private repos).
Install skills with the helper scripts.
## Communication
When listing skills, output approximately as follows, depending on the context of the user's request. If they ask about experimental skills, list from `.experimental` instead of `.curated` and label the source accordingly:
"""
Skills from {repo}:
1. skill-1
2. skill-2 (already installed)
3. ...
Which ones would you like installed?
"""
After installing a skill, tell the user it will be available on their next turn.
## Scripts
All of these scripts use network, so when running in the sandbox, request escalation when running them.
- `scripts/list-skills.py` (prints skills list with installed annotations)
- `scripts/list-skills.py --format json`
- Example (experimental list): `scripts/list-skills.py --path skills/.experimental`
- `scripts/install-skill-from-github.py --repo <owner>/<repo> --path <path/to/skill> [<path/to/skill> ...]`
- `scripts/install-skill-from-github.py --url https://github.com/<owner>/<repo>/tree/<ref>/<path>`
- Example (experimental skill): `scripts/install-skill-from-github.py --repo openai/skills --path skills/.experimental/<skill-name>`
## Behavior and Options
- Defaults to direct download for public GitHub repos.
- If download fails with auth/permission errors, falls back to git sparse checkout.
- Aborts if the destination skill directory already exists.
- Installs into `$CODEX_HOME/skills/<skill-name>` (defaults to `~/.codex/skills`).
- Multiple `--path` values install multiple skills in one run, each named from the path basename unless `--name` is supplied.
- Options: `--ref <ref>` (default `main`), `--dest <path>`, `--method auto|download|git`.
## Notes
- Curated listing is fetched from `https://github.com/openai/skills/tree/main/skills/.curated` via the GitHub API. If it is unavailable, explain the error and exit.
- Private GitHub repos can be accessed via existing git credentials or optional `GITHUB_TOKEN`/`GH_TOKEN` for download.
- Git fallback tries HTTPS first, then SSH.
- The skills at https://github.com/openai/skills/tree/main/skills/.system are preinstalled, so no need to help users install those. If they ask, just explain this. If they insist, you can download and overwrite.
- Installed annotations come from `$CODEX_HOME/skills`.

5
.codex-home/skills/.system/skill-installer/agents/openai.yaml Normal file
View file

@ -0,0 +1,5 @@
interface:
display_name: "Skill Installer"
short_description: "Install curated skills from openai/skills or other repos"
icon_small: "./assets/skill-installer-small.svg"
icon_large: "./assets/skill-installer.png"

3
.codex-home/skills/.system/skill-installer/assets/skill-installer-small.svg Normal file
View file

@ -0,0 +1,3 @@
<svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" fill="currentColor" viewBox="0 0 16 16">
<path fill="#0D0D0D" d="M2.145 3.959a2.033 2.033 0 0 1 2.022-1.824h5.966c.551 0 .997 0 1.357.029.367.03.692.093.993.246l.174.098c.397.243.72.593.932 1.01l.053.114c.116.269.168.557.194.878.03.36.03.805.03 1.357v4.3a2.365 2.365 0 0 1-2.366 2.365h-1.312a2.198 2.198 0 0 1-4.377 0H4.167A2.032 2.032 0 0 1 2.135 10.5V9.333l.004-.088A.865.865 0 0 1 3 8.468l.116-.006A1.135 1.135 0 0 0 3 6.199a.865.865 0 0 1-.865-.864V4.167l.01-.208Zm1.054 1.186a2.198 2.198 0 0 1 0 4.376v.98c0 .534.433.967.968.967H6l.089.004a.866.866 0 0 1 .776.861 1.135 1.135 0 0 0 2.27 0c0-.478.387-.865.865-.865h1.5c.719 0 1.301-.583 1.301-1.301v-4.3c0-.57 0-.964-.025-1.27a1.933 1.933 0 0 0-.09-.493L12.642 4a1.47 1.47 0 0 0-.541-.585l-.102-.056c-.126-.065-.295-.11-.596-.135a17.31 17.31 0 0 0-1.27-.025H4.167a.968.968 0 0 0-.968.968v.978Z"/>
</svg>
Side by side

After

Width:  |  Height:  |  Size: 923 B

BIN
.codex-home/skills/.system/skill-installer/assets/skill-installer.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 1.1 KiB

21
.codex-home/skills/.system/skill-installer/scripts/github_utils.py Normal file
View file

@ -0,0 +1,21 @@
#!/usr/bin/env python3
"""Shared GitHub helpers for skill install scripts."""
from __future__ import annotations
import os
import urllib.request
def github_request(url: str, user_agent: str) -> bytes:
headers = {"User-Agent": user_agent}
token = os.environ.get("GITHUB_TOKEN") or os.environ.get("GH_TOKEN")
if token:
headers["Authorization"] = f"token {token}"
req = urllib.request.Request(url, headers=headers)
with urllib.request.urlopen(req) as resp:
return resp.read()
def github_api_contents_url(repo: str, path: str, ref: str) -> str:
return f"https://api.github.com/repos/{repo}/contents/{path}?ref={ref}"

308
.codex-home/skills/.system/skill-installer/scripts/install-skill-from-github.py Normal file
View file

@ -0,0 +1,308 @@
#!/usr/bin/env python3
"""Install a skill from a GitHub repo path into $CODEX_HOME/skills."""
from __future__ import annotations
import argparse
from dataclasses import dataclass
import os
import shutil
import subprocess
import sys
import tempfile
import urllib.error
import urllib.parse
import zipfile
from github_utils import github_request
DEFAULT_REF = "main"
@dataclass
class Args:
url: str | None = None
repo: str | None = None
path: list[str] | None = None
ref: str = DEFAULT_REF
dest: str | None = None
name: str | None = None
method: str = "auto"
@dataclass
class Source:
owner: str
repo: str
ref: str
paths: list[str]
repo_url: str | None = None
class InstallError(Exception):
pass
def _codex_home() -> str:
return os.environ.get("CODEX_HOME", os.path.expanduser("~/.codex"))
def _tmp_root() -> str:
base = os.path.join(tempfile.gettempdir(), "codex")
os.makedirs(base, exist_ok=True)
return base
def _request(url: str) -> bytes:
return github_request(url, "codex-skill-install")
def _parse_github_url(url: str, default_ref: str) -> tuple[str, str, str, str | None]:
parsed = urllib.parse.urlparse(url)
if parsed.netloc != "github.com":
raise InstallError("Only GitHub URLs are supported for download mode.")
parts = [p for p in parsed.path.split("/") if p]
if len(parts) < 2:
raise InstallError("Invalid GitHub URL.")
owner, repo = parts[0], parts[1]
ref = default_ref
subpath = ""
if len(parts) > 2:
if parts[2] in ("tree", "blob"):
if len(parts) < 4:
raise InstallError("GitHub URL missing ref or path.")
ref = parts[3]
subpath = "/".join(parts[4:])
else:
subpath = "/".join(parts[2:])
return owner, repo, ref, subpath or None
def _download_repo_zip(owner: str, repo: str, ref: str, dest_dir: str) -> str:
zip_url = f"https://codeload.github.com/{owner}/{repo}/zip/{ref}"
zip_path = os.path.join(dest_dir, "repo.zip")
try:
payload = _request(zip_url)
except urllib.error.HTTPError as exc:
raise InstallError(f"Download failed: HTTP {exc.code}") from exc
with open(zip_path, "wb") as file_handle:
file_handle.write(payload)
with zipfile.ZipFile(zip_path, "r") as zip_file:
_safe_extract_zip(zip_file, dest_dir)
top_levels = {name.split("/")[0] for name in zip_file.namelist() if name}
if not top_levels:
raise InstallError("Downloaded archive was empty.")
if len(top_levels) != 1:
raise InstallError("Unexpected archive layout.")
return os.path.join(dest_dir, next(iter(top_levels)))
def _run_git(args: list[str]) -> None:
result = subprocess.run(args, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
if result.returncode != 0:
raise InstallError(result.stderr.strip() or "Git command failed.")
def _safe_extract_zip(zip_file: zipfile.ZipFile, dest_dir: str) -> None:
dest_root = os.path.realpath(dest_dir)
for info in zip_file.infolist():
extracted_path = os.path.realpath(os.path.join(dest_dir, info.filename))
if extracted_path == dest_root or extracted_path.startswith(dest_root + os.sep):
continue
raise InstallError("Archive contains files outside the destination.")
zip_file.extractall(dest_dir)
def _validate_relative_path(path: str) -> None:
if os.path.isabs(path) or os.path.normpath(path).startswith(".."):
raise InstallError("Skill path must be a relative path inside the repo.")
def _validate_skill_name(name: str) -> None:
altsep = os.path.altsep
if not name or os.path.sep in name or (altsep and altsep in name):
raise InstallError("Skill name must be a single path segment.")
if name in (".", ".."):
raise InstallError("Invalid skill name.")
def _git_sparse_checkout(repo_url: str, ref: str, paths: list[str], dest_dir: str) -> str:
repo_dir = os.path.join(dest_dir, "repo")
clone_cmd = [
"git",
"clone",
"--filter=blob:none",
"--depth",
"1",
"--sparse",
"--single-branch",
"--branch",
ref,
repo_url,
repo_dir,
]
try:
_run_git(clone_cmd)
except InstallError:
_run_git(
[
"git",
"clone",
"--filter=blob:none",
"--depth",
"1",
"--sparse",
"--single-branch",
repo_url,
repo_dir,
]
)
_run_git(["git", "-C", repo_dir, "sparse-checkout", "set", *paths])
_run_git(["git", "-C", repo_dir, "checkout", ref])
return repo_dir
def _validate_skill(path: str) -> None:
if not os.path.isdir(path):
raise InstallError(f"Skill path not found: {path}")
skill_md = os.path.join(path, "SKILL.md")
if not os.path.isfile(skill_md):
raise InstallError("SKILL.md not found in selected skill directory.")
def _copy_skill(src: str, dest_dir: str) -> None:
os.makedirs(os.path.dirname(dest_dir), exist_ok=True)
if os.path.exists(dest_dir):
raise InstallError(f"Destination already exists: {dest_dir}")
shutil.copytree(src, dest_dir)
def _build_repo_url(owner: str, repo: str) -> str:
return f"https://github.com/{owner}/{repo}.git"
def _build_repo_ssh(owner: str, repo: str) -> str:
return f"git@github.com:{owner}/{repo}.git"
def _prepare_repo(source: Source, method: str, tmp_dir: str) -> str:
if method in ("download", "auto"):
try:
return _download_repo_zip(source.owner, source.repo, source.ref, tmp_dir)
except InstallError as exc:
if method == "download":
raise
err_msg = str(exc)
if "HTTP 401" in err_msg or "HTTP 403" in err_msg or "HTTP 404" in err_msg:
pass
else:
raise
if method in ("git", "auto"):
repo_url = source.repo_url or _build_repo_url(source.owner, source.repo)
try:
return _git_sparse_checkout(repo_url, source.ref, source.paths, tmp_dir)
except InstallError:
repo_url = _build_repo_ssh(source.owner, source.repo)
return _git_sparse_checkout(repo_url, source.ref, source.paths, tmp_dir)
raise InstallError("Unsupported method.")
def _resolve_source(args: Args) -> Source:
if args.url:
owner, repo, ref, url_path = _parse_github_url(args.url, args.ref)
if args.path is not None:
paths = list(args.path)
elif url_path:
paths = [url_path]
else:
paths = []
if not paths:
raise InstallError("Missing --path for GitHub URL.")
return Source(owner=owner, repo=repo, ref=ref, paths=paths)
if not args.repo:
raise InstallError("Provide --repo or --url.")
if "://" in args.repo:
return _resolve_source(
Args(url=args.repo, repo=None, path=args.path, ref=args.ref)
)
repo_parts = [p for p in args.repo.split("/") if p]
if len(repo_parts) != 2:
raise InstallError("--repo must be in owner/repo format.")
if not args.path:
raise InstallError("Missing --path for --repo.")
paths = list(args.path)
return Source(
owner=repo_parts[0],
repo=repo_parts[1],
ref=args.ref,
paths=paths,
)
def _default_dest() -> str:
return os.path.join(_codex_home(), "skills")
def _parse_args(argv: list[str]) -> Args:
parser = argparse.ArgumentParser(description="Install a skill from GitHub.")
parser.add_argument("--repo", help="owner/repo")
parser.add_argument("--url", help="https://github.com/owner/repo[/tree/ref/path]")
parser.add_argument(
"--path",
nargs="+",
help="Path(s) to skill(s) inside repo",
)
parser.add_argument("--ref", default=DEFAULT_REF)
parser.add_argument("--dest", help="Destination skills directory")
parser.add_argument(
"--name", help="Destination skill name (defaults to basename of path)"
)
parser.add_argument(
"--method",
choices=["auto", "download", "git"],
default="auto",
)
return parser.parse_args(argv, namespace=Args())
def main(argv: list[str]) -> int:
args = _parse_args(argv)
try:
source = _resolve_source(args)
source.ref = source.ref or args.ref
if not source.paths:
raise InstallError("No skill paths provided.")
for path in source.paths:
_validate_relative_path(path)
dest_root = args.dest or _default_dest()
tmp_dir = tempfile.mkdtemp(prefix="skill-install-", dir=_tmp_root())
try:
repo_root = _prepare_repo(source, args.method, tmp_dir)
installed = []
for path in source.paths:
skill_name = args.name if len(source.paths) == 1 else None
skill_name = skill_name or os.path.basename(path.rstrip("/"))
_validate_skill_name(skill_name)
if not skill_name:
raise InstallError("Unable to derive skill name.")
dest_dir = os.path.join(dest_root, skill_name)
if os.path.exists(dest_dir):
raise InstallError(f"Destination already exists: {dest_dir}")
skill_src = os.path.join(repo_root, path)
_validate_skill(skill_src)
_copy_skill(skill_src, dest_dir)
installed.append((skill_name, dest_dir))
finally:
if os.path.isdir(tmp_dir):
shutil.rmtree(tmp_dir, ignore_errors=True)
for skill_name, dest_dir in installed:
print(f"Installed {skill_name} to {dest_dir}")
return 0
except InstallError as exc:
print(f"Error: {exc}", file=sys.stderr)
return 1
if __name__ == "__main__":
raise SystemExit(main(sys.argv[1:]))

107
.codex-home/skills/.system/skill-installer/scripts/list-skills.py Normal file
View file

@ -0,0 +1,107 @@
#!/usr/bin/env python3
"""List skills from a GitHub repo path."""
from __future__ import annotations
import argparse
import json
import os
import sys
import urllib.error
from github_utils import github_api_contents_url, github_request
DEFAULT_REPO = "openai/skills"
DEFAULT_PATH = "skills/.curated"
DEFAULT_REF = "main"
class ListError(Exception):
pass
class Args(argparse.Namespace):
repo: str
path: str
ref: str
format: str
def _request(url: str) -> bytes:
return github_request(url, "codex-skill-list")
def _codex_home() -> str:
return os.environ.get("CODEX_HOME", os.path.expanduser("~/.codex"))
def _installed_skills() -> set[str]:
root = os.path.join(_codex_home(), "skills")
if not os.path.isdir(root):
return set()
entries = set()
for name in os.listdir(root):
path = os.path.join(root, name)
if os.path.isdir(path):
entries.add(name)
return entries
def _list_skills(repo: str, path: str, ref: str) -> list[str]:
api_url = github_api_contents_url(repo, path, ref)
try:
payload = _request(api_url)
except urllib.error.HTTPError as exc:
if exc.code == 404:
raise ListError(
"Skills path not found: "
f"https://github.com/{repo}/tree/{ref}/{path}"
) from exc
raise ListError(f"Failed to fetch skills: HTTP {exc.code}") from exc
data = json.loads(payload.decode("utf-8"))
if not isinstance(data, list):
raise ListError("Unexpected skills listing response.")
skills = [item["name"] for item in data if item.get("type") == "dir"]
return sorted(skills)
def _parse_args(argv: list[str]) -> Args:
parser = argparse.ArgumentParser(description="List skills.")
parser.add_argument("--repo", default=DEFAULT_REPO)
parser.add_argument(
"--path",
default=DEFAULT_PATH,
help="Repo path to list (default: skills/.curated)",
)
parser.add_argument("--ref", default=DEFAULT_REF)
parser.add_argument(
"--format",
choices=["text", "json"],
default="text",
help="Output format",
)
return parser.parse_args(argv, namespace=Args())
def main(argv: list[str]) -> int:
args = _parse_args(argv)
try:
skills = _list_skills(args.repo, args.path, args.ref)
installed = _installed_skills()
if args.format == "json":
payload = [
{"name": name, "installed": name in installed} for name in skills
]
print(json.dumps(payload))
else:
for idx, name in enumerate(skills, start=1):
suffix = " (already installed)" if name in installed else ""
print(f"{idx}. {name}{suffix}")
return 0
except ListError as exc:
print(f"Error: {exc}", file=sys.stderr)
return 1
if __name__ == "__main__":
raise SystemExit(main(sys.argv[1:]))