Describe Image Skill — Development skill for Claude Code

describe-image

Let non-vision models "see" images: hand the image to a separate vision model, get its description as text, and let your host model answer from that text.

让没有视觉能力的模型也能"看懂"图片：把图片交给一个独立的视觉模型转述成文字，再由你的宿主模型基于这段文字回答。

预览 / Preview

目前实测支持 claude code、opencode，并内置 Codex、Gemini CLI 的会话记录解析与记忆写入；其他宿主可在 `config.json` 里指定扫描路径与记忆文件。

If your model does not accept images (e.g. DeepSeek, some local models), pasted screenshots only appear to you as `[Image: ...]` / `[Unsupported Image]` placeholders — you never see the pixels. This skill is built for exactly that case: it forwards the image to a vision model (SiliconFlow / OpenAI / Ollama / any OpenAI-compatible endpoint) and returns the recognition result as plain text that you can reason over.

如果你的模型不接收图片（例如 DeepSeek、部分本地模型），用户在对话里粘贴截图时，你只能看到 `[Image: ...]`、`[Unsupported Image]` 这类占位符，看不到像素。这个技能就是为这种情况准备的：它把图片交给一个视觉模型（SiliconFlow / OpenAI / Ollama / 任意 OpenAI 兼容端点），把识别结果以纯文本返回，你再用这段文本回答用户。

Use cases: screenshot content recognition, transcribing text inside images, UI description, interpreting charts / documents / photos, and any "what is in this image" question.

适用场景：截图内容识别、图片里的文字转录、UI 界面描述、图表/文档/照片解读，以及一切"图片里有什么"的问题。

Why / 为什么需要它

• Your host model has no vision capability (DeepSeek, certain local models, etc.).
• 你的宿主模型没有视觉能力（DeepSeek、某些本地模型等）。
• The user pasted an image, but all you get is a placeholder; the actual pixels only live in the session transcript.
• 用户粘贴了图片，但你能拿到的只是占位符文本，图片本体只在底层会话记录里。
• You do not want to switch models over a single image — you just want to call a vision model on demand.
• 你不想因为一张图换掉整个模型，只想在需要的时候临时调用一个视觉模型。

This skill does not require you to change models or to have a hos