AI Agent Hub
Back to plugins
🖥️

dsh-wechat-pro

Client Updated 2026.08.18

Run the following command in DeepSeek Harness:

dsh plugin install bwhite55/dsh-wechat-pro

Paste the following prompt into your AI chat to install this plugin:

Install by running dsh plugin install bwhite55/dsh-wechat-pro in your terminal; full source code is available at https://github.com/bwhite55/dsh-wechat-pro .

About this plugin

DSH (DeepSeek Harness) is a locally-run AI coding assistant, but it lives on your desktop. Dropped a connection during a train ride, stuck in a meeting, or simply on the couch and want to check on a long-running task or approve a risky tool call? dsh-wechat-pro bridges DSH into personal WeChat so you can chat with your AI from your phone. No public server, no extra daemon, no OpenClaw. Scan a QR code and you are connected.

The plugin runs inside the dsh web process and talks to WeChat over Tencent's official iLink channel, with credentials that auto-refresh every 24 hours. Sessions you open in WeChat are real DSH sessions, visible in the Web session list and continuable across devices. Commands such as /model, /thinking, /workspace, and native harness commands like /plan and /permission work directly in chat. High-risk tool-call approvals are pushed to WeChat and resolved with a simple /yes or /no. Streaming output is pushed in real time, /send delivers host files to WeChat, incoming media (images, files, videos, voice) is automatically downloaded and decrypted, and the agent can proactively push messages via the weixin_send tool.

If you already run DSH locally and want a pocket-sized client to check on progress or approve actions between meetings, from the couch, or during a commute, this is a zero-infrastructure fit. All data stays on your machine, the control API binds to loopback only, and the allowFrom whitelist keeps access limited to your own WeChat ID.

Use Cases

  • Approve high-risk tool calls from your phone during a meeting
  • Check streaming output and final results of a long DSH task while commuting
  • Switch workspace and model in chat without going back to your desk

Best For

  • Developers already running DSH locally
  • Users who want remote access without a public server or extra daemon
  • Teams that need to approve and monitor AI coding tasks from mobile devices