codex-image-generator
/codex-image-generator
Default image generation path. Generate or edit raster images (hero images, illustrations, textures, product shots, mockups, transparent cutouts) through the Codex CLI's built-in image_gen tool, run in a subagent. Use whenever the user asks for an image, picture, illustration, or visual asset and does not name a provider. Not for SVG, icons, or diagrams that belong in code.
- Author
- stevenmays ↗
- Repository
- dotfiles ↗
- Licence
- MIT
- Last repository change
- 21.10.2026
Assessment and security
- Files in this package
- 2
- Scripts
- 1
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 21.10.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
- 1 scripts in this package. Review them before installing, as Claude may execute them.
Scripts: scripts/generate.sh
What it does and when to use it
Default image generation path. Generate or edit raster images (hero images, illustrations, textures, product shots, mockups, transparent cutouts) through the Codex CLI's built-in image_gen tool, run in a subagent. Use whenever the user asks for an image, picture, illustration, or visual asset and does not name a provider. Not for SVG, icons, or diagrams that belong in code.
Original author description
Default image generation path. Generate or edit raster images (hero images, illustrations, textures, product shots, mockups, transparent cutouts) through the Codex CLI's built-in image_gen tool, run in a subagent. Use whenever the user asks for an image, picture, illustration, or visual asset and does not name a provider. Not for SVG, icons, or diagrams that belong in code.
How it works
- Run it
- Rules
SKILL.md preview
---
name: codex-image-generator
description: Default image generation path. Generate or edit raster images (hero images, illustrations, textures, product shots, mockups, transparent cutouts) through the Codex CLI's built-in image_gen tool, run in a subagent. Use whenever the user asks for an image, picture, illustration, or visual asset and does not name a provider. Not for SVG, icons, or diagrams that belong in code.
---
# Codex image generator
Codex CLI ships an `image_gen` tool and a bundled `imagegen` skill that shapes prompts, saves under `$CODEX_HOME/generated_images/`, and copies the result to a requested path. This skill wraps one `codex exec` call around that tool and verifies the file. No API key is needed; Codex uses its own login.
## Run it
Delegate the run to a `general-purpose` subagent pinned to `model: "sonnet"`. A Codex run takes about 50 seconds and the result is a file check, so the session must not block on it. Give the subagent the exact command and the verification rule below.
```bash
"${CLAUDE_PLUGIN_ROOT}/skills/codex-image-generator/scripts/generate.sh" \
--prompt "<image description>" \
--output <path>.png \
[--reference <image>]... [--model gpt-6-astra] [--open]
```
- `--prompt`: describe scene, subject, style, composition, and lighting. Quote exact in-image text verbatim. Ask for a transparent background when the asset must sit on the page's color.
- `--output`: the final PNG path. The script creates the parent directory and rejects a run that leaves no image there.
- `--reference`: attaches an image to the Codex prompt. Use it for style guidance or to edit an existing image; say in the prompt which role each reference plays and what must stay unchanged. Edits drift: in testing a sky-only change also altered style, aspect ratio, and background detail, so list every invariant and check the result.
- `--model`: defaults to `gpt-5.6-sol`. `gpt-6-astra` produces the same image quality at higher cost, so use it only when the user asks.
- `--open`: opens the verified image in Preview. Use it for one-off requests so the user sees the result. Skip it for assets that code consumes.
The script prints the absolute path, then the `file` description, byte size, elapsed seconds, and model. Exit code 5 means Codex finished without leaving an image at the path; the log path in stderr shows what Codex did.
## Rules
- One asset per call. For variants or a set, run the script once per asset with distinct prompts.
- Never overwrite an existing asset the user did not ask to replace. Use a sibling name such as `hero-v2.png`.
- Report the saved path and the final prompt. Do not paste the image into chat.
- If `codex` is missing or not logged in, say so and stop.
Files in this package (2)
- SKILL.md 3 KB
- scripts/generate.sh 2 KB
Versions
The database has no other skills named codex-image-generator.
Database snapshot: 13.11.2026 · package version 025ef9cb4b1e (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 100codex-image
Generate or edit raster images (photos, illustrations, mockups, hero images, product shots, bitmap icons, transparent cutouts) with OpenAI GPT Image 2 by triggering the local Codex CLI, billed to the user's ChatGPT subscription — no API key needed. Use when the user says… - 100gemini-image-generator
Generate or edit raster images with Google's Nano Banana model through the Antigravity CLI's built-in generate_image tool, run in a subagent. Use when the user names Gemini, Nano Banana, or Antigravity. Not for SVG, icons, or code diagrams. The default image path is codex-image-generator. - 100image
Generates raster images (icons, illustrations, textures) via the logged-in Codex CLI image tool. Use when an agent needs a real image but has no native image generation. - 100codex-imagegen
Generate or edit raster images (photos, illustrations, mockups, logos, product shots, game/UI assets, textures, infographics) using Codex's built-in image_gen tool — driven headlessly through `codex exec`. Billed via the user's ChatGPT/Codex login, so it needs NO OpenAI API key.… - 95imagegenkit
Generate or edit images with the local `imagegenkit` command (ImageGenKit). Use it whenever the user asks to create, generate, render, mock up or edit an image, or a task needs a raster image asset (photos, product shots, thumbnails, mockups, icons, illustrations, backgrounds,… - 100imagegen
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 the active client should create a new image, transform an existing image, or derive visual…