AI Agent Hub
Back to plugins
🤖

dsh-lark

Model Inference Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install sugarforever/dsh-lark

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

Install the dsh-lark plugin from https://github.com/sugarforever/dsh-lark in DeepSeek Harness to connect Feishu or Lark.

About this plugin

dsh-lark brings DeepSeek Harness agents directly into Feishu or Lark. It addresses a practical problem: how to use an already-configured Harness agent inside enterprise chat without exposing a public server, domain, or webhook callback. The plugin uses the official Feishu/Lark SDK over a WebSocket long connection, so it can run locally, on an internal network, or in a normal development environment.

For conversations, the plugin maps Feishu or Lark chats to Harness sessions. Direct chats and normal group chats reuse a session by chat, while topic groups keep separate sessions per thread. Replies are linked to the original message and stay in the same thread. Group chats require @mention by default, and you can use allowlists or channel-specific model settings when needed.

This plugin is a good fit for developers or internal platform teams who already use DeepSeek Harness and want to expose agents in Feishu or Lark. It is especially useful for teams that need model inference, tool access, session storage, and access control inside enterprise messaging without maintaining a public webhook service.

Use Cases

  • Use a Harness Agent in Feishu direct messages
  • Ask questions by mentioning the bot in group chats
  • Keep topic-based conversations in threaded groups

Best For

  • Harness users who want to connect Feishu or Lark
  • Teams that need an internal bot for questions
  • Developers who do not want public webhook callbacks