菜单
Revise a Cowart canvas image from an annotation screenshot using Vetta edit_image, then place the result with Cowart MCP. Use when the user submits a Cowart annotation edit or asks to clean up an annotated screenshot into a revised bitmap beside the original.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/openvetta/open-vetta --skill cowart-image-edit命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
基于 SOC 职业分类
| name | cowart-image-edit |
| description | Revise a Cowart canvas image from an annotation screenshot using Vetta edit_image, then place the result with Cowart MCP. Use when the user submits a Cowart annotation edit or asks to clean up an annotated screenshot into a revised bitmap beside the original. |
| Intent | Tool |
|---|---|
| Modify existing image / apply annotation marks | edit_image |
| Brand-new image with no source to revise | generate_image (use cowart-image-gen skill) |
Do not use an imagegen skill. After edit_image writes a file, place it with Cowart MCP insert_cowart_image.
Obtain the annotation screenshot (user attachment or Cowart-exported capture path from the follow-up prompt).
Call edit_image with that screenshot (and original image if available) so the model applies annotated intent and removes annotation chrome (arrows, scribbles, labels) unless the user wants them kept.
Place the result with MCP insert_cowart_image beside the original when possible (placement: "right"); do not delete or move the original unless asked.
Use projectDir = active user workspace for all Cowart MCP calls.
edit_image for the bitmap step, then Cowart MCP for placement.get_cowart_selection can help locate the target image.