Back to directory
AI & Automationdamnji/vision-skill
vision-skill
给不支持多模态的 AI 模型（如 DeepSeek）提供识图能力——把图片发给云端视觉模型，返回文字描述。支持任意 OpenAI 兼容接口与 Anthropic 原生接口、主备双模型自动切换、BMP 智能路由、报错自动翻译。
Suggested install command
npx skills add damnji/vision-skill/vision-skillAlways inspect the linked repository and skill instructions before running commands. Skills are instructions; permissions and execution still matter.
Compatibility
Agent support matrix
3 supported
| Agent | Status |
|---|---|
| Claude Code | Supported |
| OpenCode | Not listed |
| Cursor | Supported |
| MCP | Not listed |
| GitHub Copilot | Not listed |
| Windsurf |