Repolode
Submit
xmanrui

xmanrui/dsh-im

A plugin that connects IM bots (Feishu, WeChat, DingTalk, WeCom, QQ, Slack, Telegram, Discord, WhatsApp) to the DeepSeek Harness via QR code or credentials.

View on GitHub
Stars
2,274
Forks
237
Open issues
60
Last commit
9 hours ago

What is it?

dsh-im is a plugin for DeepSeek Harness that lets you connect chat bots from multiple IM platforms to a local Harness instance. Bots can be bound by scanning a QR code, using an app manifest, or entering existing bot credentials, and each channel supports several bots that keep their own workspace, model, Agent Preset and session mapping.

The built-in channels cover Feishu, WeChat, DingTalk, WeCom (bot and self-built app), QQ, Slack, Telegram, Discord, WhatsApp and iMessage on macOS, with Matrix available as an experimental channel. Each channel uses its platform's own long-connection or polling mechanism, and the plugin supports streaming replies, progress indicators, image input, result file return, approval handling and group chat history where the platform allows it.

All credentials are submitted only to the local Harness Host and stored in protected storage; the browser receives only QR codes, redacted status and the access settings the user saves.

Key features

  • 01Connects bots from nine-plus IM platforms to a local DeepSeek Harness.
  • 02Supports QR-code onboarding, app manifests, or manual credential entry.
  • 03Manages multiple bots per channel, each with its own workspace, model and session mapping.
  • 04Streams replies and shows tool progress where the platform supports it.
  • 05Accepts image input and returns result files as native attachments.
  • 06Supports remote approval of tool actions by replying in chat.

Use cases

  • Chatting with a local coding or research agent from a team's Feishu or DingTalk group.
  • Reaching your Harness from Telegram or WhatsApp while away from the desktop.
  • Running one Harness behind several customer-facing bots on different platforms.

Quick start

  1. 1

    Install the plugin for DSH Desktop from npm

    dsh plugin --profile desktop add -w @xmanrui/dsh-im
  2. 2

    Restart DSH Desktop, then open Settings and go to IM机器人

  3. 3

    In the channel page, connect a bot by scanning a QR code or entering credentials

Things to know

  • Requires the DeepSeek Harness (DSH) to run, so it is only useful inside that ecosystem.
  • iMessage works only on macOS through Messages.app.
  • Matrix end-to-end encryption is experimental and not suitable for confidential use.

Topics

#chatbot#im-bot#deepseek#ai-agents#feishu#telegram-bot#discord-bot#slack-bot#cordis#deepseek-harness

FAQ

What is dsh-im?
A plugin that connects IM bots (Feishu, WeChat, DingTalk, WeCom, QQ, Slack, Telegram, Discord, WhatsApp) to the DeepSeek Harness via QR code or credentials. dsh-im is a plugin for DeepSeek Harness that lets you connect chat bots from multiple IM platforms to a local Harness instance. Bots can be bound by scanning a QR code, using an app manifest, or entering existing bot credentials, and each channel supports several bots that keep their own workspace, model, Agent Preset and session mapping.
Is dsh-im free to use? What license does it have?
dsh-im is open source under the MIT license. Check the license terms before using it in commercial products.
Is dsh-im still maintained?
Actively maintained. The last commit to dsh-im was on Oct 21, 2026.
What can I use dsh-im for?
Chatting with a local coding or research agent from a team's Feishu or DingTalk group. Reaching your Harness from Telegram or WhatsApp while away from the desktop. Running one Harness behind several customer-facing bots on different platforms.
How do I install dsh-im?
Install the plugin for DSH Desktop from npm: dsh plugin --profile desktop add -w @xmanrui/dsh-im Restart DSH Desktop, then open Settings and go to IM机器人 In the channel page, connect a bot by scanning a QR code or entering credentials
Where can I get dsh-im?
The source code and installation instructions are on GitHub: https://github.com/xmanrui/dsh-im

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects