Codex Image Generation
Generate, edit, and prepare practical raster image assets directly within a project workflow.
The documentation separates built-in and CLI modes, requires confirmation for the transparency downgrade, avoids overwriting files, handles API keys cautiously, and discusses network-approval risk. It does not fully disclose image/prompt data flows, retention or privacy behavior, built-in-tool external effects, or a general rollback mechanism, so 7 points are deducted.
The script and docs provide parameter validation, missing-dependency guidance, output-collision protection, batch limits, and some transient-error retries; the main path is plausible from static inspection. However, there is no committed test suite or execution evidence, and built-in-tool failure coverage and recovery remain limited, so the score is capped conservatively.
Use cases, generate/edit distinctions, non-fit cases, trigger rules, and output handling are clearly described. Chinese-language support is not documented, and the CLI depends on OpenAI network services without evidence of mainland-China reachability or a practical alternative, so 5 points are deducted.
The package has strong information architecture, progressive disclosure, dependency notes, parameters, examples, limitations, and cross-references. It lacks an intrinsic version, changelog, explicit maintenance owner, and update path; license and governance are mainly supplied as external metadata, so 4 points are deducted.
The skill covers generation, editing, batching, transparency workflows, and project-local artifact handling, with a clear core benefit. Static files do not verify output quality, text accuracy, or iteration outcomes, so limited manual review remains necessary and 1 point is deducted.
The SKILL.md, reference documents, and CLI source provide auditable rules, parameters, and diagnostics, making some claims source-checkable. There is no committed test suite, CI coverage, or third-party execution record, so the static evidence supports only a low score.
- Image generation depends on an external model service; handling, retention, and privacy boundaries for prompts and input images are not explicitly documented.
- CLI mode requires OPENAI_API_KEY, Python dependencies, and network access; mainland-China users may face service reachability or network-approval issues.
- No tests or CI evidence cover the key paths, so output quality, text accuracy, transparency edges, and edit invariants require manual verification.
What it does & when to use it
This skill is designed for Codex image generation and editing, covering photos, illustrations, textures, sprites, mockups, and transparent-background assets. Its preferred path uses the built-in image_gen tool and does not require OPENAI_API_KEY; the CLI fallback requires an API key and additional dependencies. It supports new images, reference-guided generation, edits to visible images, and repeated asset variants. It is not intended for deterministic SVG, vector, HTML/CSS, canvas, or other code-native graphics work.
Generates raster images from prompts; edits images visible in the conversation; inspects a local edit target with view_image before built-in editing; handles simple transparency by generating a flat chroma-key background and running remove_chroma_key.py to produce PNG/WebP output with alpha; runs scripts/image_gen.py with generate, edit, or generate-batch in explicit CLI fallback mode; validates outputs and iterates with one targeted change at a time.
- A web or product team needs a landing-page hero, product shot, or marketing image.
- A game developer needs concept art, sprites, or other bitmap assets.
- A designer wants style or composition variants based on reference images.
- A user needs background replacement, object removal, or lighting and weather changes while preserving the rest of an image.
- A team needs UI mockups, wireframes, infographics, or scientific-educational visuals.
Pros & cons
- Covers generation, editing, reference-based variants, batches, and simple transparent-background processing.
- The preferred built-in mode does not require an API key.
- Provides structured prompting, use-case taxonomy, edit invariants, and output-validation guidance.
- Clearly separates raster generation from SVG, HTML/CSS, canvas, and other code-native workflows.
- It contains Codex-specific tools, paths, and environment conventions, so other clients require adaptation.
- CLI fallback requires OPENAI_API_KEY, Python, the openai package, and Pillow.
- Complex transparent subjects such as hair, smoke, glass, liquids, or reflective materials require confirmation before using the true-transparency fallback.
- The source does not document a standalone skill installer, test coverage, or cross-platform validation.
How to install
The repository bundles this skill with 19 others at codex-rs/skills/src/assets/samples/imagegen/SKILL.md. The README documents Codex CLI installation but does not document a standalone installer for this skill. Following the repository instructions, install Codex CLI with curl -fsSL https://chatgpt.com/codex/install.sh | sh on macOS/Linux, or powershell -ExecutionPolicy ByPass -c "irm https://chatgpt.com/codex/install.ps1 | iex" on Windows. npm install -g @openai/codex and brew install --cask codex are also documented options.
How to use
In Codex, provide a concrete image request such as: “Generate a minimal ceramic coffee mug product image for a website hero, using clean product photography, with usable negative space for copy and no logo, text, or watermark.” Normal generation and editing use the built-in image_gen path. Choose scripts/image_gen.py only when explicitly requesting CLI/API/model control; the CLI fallback requires OPENAI_API_KEY.