✨ Why

DSH's default text-only model (e.g. deepseek-v4-flash) can't see images. This plugin gives it eyes via a small, free vision model — no local GPU, no image-editing, no changes to your model.

🚀 Features

• see_image tool — pass any image path + a question; get a text description back.
• Vision-model-agnostic — any OpenAI-compatible /chat/completions endpoint (Zhipu GLM-4V-Flash by default, or SiliconFlow / Qwen2.5-VL, …).
• Sandbox-aware — reads through ctx.fs, respecting DSH's sandbox / observation policy.
• Bonus paste-to-text — a patch script that lets you paste an image straight into the chat and auto-convert it to text.

📚 Contents

• How it works
• Install (DSH web profile)
• Configuration
• Uninstall / rollback
• Implementation notes
• Regression test
• Field notes
• Paste-to-text relay (bonus)
• License

🔍 How it works

You: "Look at this image" ──►  Text-only model (no vision)
                                  │ calls see_image(path, question)
                                  ▼
                             This plugin (Host plane)
                                  │ 1. ctx.fs resolves & reads the image (sandbox/observation policy aware)
                                  │ 2. encodes it as a base64 data URL
                                  │ 3. POST {baseURL}/chat/completions (OpenAI-compatible)
                                  ▼
                             Vision model (GLM-4V-Flash)
                                  │ text description
                                  ▼
                             Text-only model ──► reports to you

📦 Install (DSH web profile)

1. Copy the plugin into your profile directory, e.g. $DSH_HOME/profiles/web/plugins/dsh-tool-see-image/ ($DSH_HOME is usually ~/.dsh).

2. Declare the dependency in $DSH_HOME/profiles/web/package.json:

   "dsh-tool-see-image": "file:plugins/dsh-tool-see-image"

   Then run pnpm install (creates a junction to the source under profiles/node_modules).

3. Compose it into the profile in $DSH_HOME/profiles/web/cordis.patch.yml:

   - insert:
       - id: tool-see-image
         name: 'dsh-tool-see-image'
         config:
           baseURL: https://open.bigmodel.cn/api/paas/v4
           apiKeyEnv: ZHIPU_API_KEY
           model: glm-4v-flash

4. Set your API key: create one at the Zhipu (bigmodel) console (format id.secret), then set the environment variable (Windows example):

   setx ZHIPU_API_KEY "your-key"

   Restart your terminal, then restart dsh web (the web profile does not hot-reload patch layers yet — tested).

5. Verify: in a new session the see_image tool should appear. Try it:

   Use see_image to look at path/to/your/image.png

⚙️ Configuration (tool-see-image line in cordis.patch.yml)

To use a different vision API, change these three keys, e.g. SiliconFlow:

config:
  baseURL: https://api.siliconflow.cn/v1
  apiKeyEnv: SILICONFLOW_API_KEY
  model: Qwen/Qwen2.5-VL-32B-Instruct

🗑️ Uninstall / rollback

1. Remove the - insert: ... tool-see-image ... block from cordis.patch.yml;
2. Remove the junction: Remove-Item profiles\node_modules\dsh-tool-see-image;
3. Remove that line from profiles/web/package.json dependencies;
4. Restart dsh web.

🧠 Implementation notes (for plugin developers)

• Exports { name, inject, Config, apply }, same shape as every DSH tool plugin;
• inject: ["tools", "fs"] — the tool registry and the sandboxed file service are both Host-global services;
• Registered at the global layer, so every session sees it (same as TUI-mode host tool rows);
• Reads files through ctx.fs (sandbox/observation policy applied), never raw node:fs;
• Parameter schema uses the DSH-specific format: required: true for required, omit the required key for optional (required: false is rejected by defineTool);
• Network request carries a timeout and exec.signal cancellation; errors are model-readable.

🧪 Regression test

test/mount-test.mjs mounts this plugin line under a real Cordis Loader (timer + system-prompt + tools + this plugin) and asserts that see_image lands in the tool registry. Run:

$env:DSH_CHECKOUT="<your dsh install root, containing node_modules/@deepseek-ai>"
node test/mount-test.mjs

Expected output: tools.schemas() 含 see_image: true and === MOUNT TEST PASS ===. The script temporarily links the plugin into the checkout's node_modules (Windows junction / other-platform symlink) and cleans up afterwards — no hardcoded local paths.

📝 Field notes

• 2026-08-13: real key + triz-workflow.png (a DSH Web GUI screenshot) → HTTP 200 in ~6.8s, correctly read the UI text (search box / MCP settings / Fetch / Filesystem / Sequential-Thinking).
• Pitfall: glm-4v-flash's max_tokens cap is 1024 (the default of 2048 caused a 400; the default has been fixed).

📋 Paste-to-text relay (bonus)

Beyond the see_image tool (which reads an image by path), this repo ships a patch script that lets you paste images straight into the chat box and have them auto-converted to text:

$env:DSH_CHECKOUT="<your dsh install root, containing node_modules/@deepseek-ai>"
node scripts/patch-dsh-image-relay.mjs          # apply (idempotent, auto-backup)
node scripts/patch-dsh-image-relay.mjs --check  # status
node scripts/patch-dsh-image-relay.mjs --revert # rollback
pm2 restart dsh-web                              # then hard-refresh the browser

It patches three DSH packages (host-apiproxy, llm-deepseek, client-ui) so that:

• the UI shows the pasted image (the hidden text block is filtered out);
• DeepSeek receives a 【图片：...】 description from GLM-4V-Flash instead of raw pixels;
• repeated images hit a local cache (~/.dsh/cache/image-relay/), and failures degrade gracefully within 8s.

Requires ZHIPU_API_KEY. The image cache itself is portable and stays under ~/.dsh/cache/image-relay/; only the DSH checkout must be supplied explicitly. Re-run the script after any npx dsh upgrade — the patch is lost when the npm cache is refreshed.

📄 License

MIT