MuseBot – Multi‑platform LLM chatbot

What it is – MuseBot is an open‑source Go application that turns a variety of communication platforms (Telegram, Discord, Slack, Lark/Feishu, DingTalk, Enterprise WeChat, QQ, WeChat) into AI‑powered chat bots. It forwards user messages to a chosen large‑language‑model (LLM) API (OpenAI, DeepSeek, Gemini, OpenRouter, OrcaRouter, 302.AI, etc.) and streams the model’s reply back to the user.

---

Core features (as listed in the README)

• AI responses via many LLM providers (text, image, video, TTS where supported).
• Streaming output – replies are sent token‑by‑token for a smoother chat experience.
• Multimodal – can send images to the model and receive image generation, voice input, and function‑call handling.
• RAG support – attach external context to prompts.
• Admin platform for managing bots, registration service for auto‑discovering instances, metrics, and cron jobs to trigger LLM calls.
• Easy deployment – run locally with go run or use the provided Docker images (Docker Hub and Aliyun registry).
• Extensive platform coverage – Telegram, Discord, Slack (socket mode & events API), Lark, DingTalk, Enterprise WeChat, QQ (ONEBOT), WeChat Official Account.

---

Quick start (from the README)

1. Clone & build

   git clone https://github.com/yincongcyincong/MuseBot.git
   cd MuseBot
   go mod tidy   # fetch Go dependencies
   

2. Set required env vars (example for Telegram + DeepSeek)

   export TELEGRAM_BOT_TOKEN="<your‑telegram‑bot‑token>"
   export DEEPSEEK_TOKEN="<your‑deepseek‑api‑key>"
   

3. Run locally

   go run main.go -telegram_bot_token=$TELEGRAM_BOT_TOKEN -deepseek_token=$DEEPSEEK_TOKEN
   

4. Or run with Docker

   docker pull jackyin0822/musebot:latest
   docker run -d \
     -v /home/user/data:/app/data \
     -e TELEGRAM_BOT_TOKEN="$TELEGRAM_BOT_TOKEN" \
     -e DEEPSEEK_TOKEN="$DEEPSEEK_TOKEN" \
     -p 36060:36060 \
     --name my-musebot jackyin0822/musebot:latest
   

   (Aliyun image is also available.)

---

Configuration highlights

Variable	Purpose	Typical value
TYPE	Which LLM provider to use (e.g., deepseek, openai, gemini, openrouter…)	deepseek
MEDIA_TYPE	Source for image/video/TTS generation	vol (Volcano Engine)
DB_TYPE	Persistence backend – sqlite3 (default) or mysql
LANG	UI language – en or zh
TOKEN_PER_USER	Per‑user token limit (0 = unlimited)
MAX_USER_CHAT	Max concurrent chats per user
LLM_PROXY / ROBOT_PROXY	Optional HTTP proxy for LLM or bot traffic

All other platform‑specific tokens (Telegram, Discord, Slack, Lark, DingTalk, WeChat, QQ, etc.) are set via similarly named environment variables.

---

Documentation & demos

• Video demos – YouTube links for a generic setup and for each LLM provider (deepseek, gemini, chatgpt).
• Platform docs – Markdown files under static/doc/ cover Discord, Slack, Lark, DingTalk, WeChat, QQ, image handling, audio handling, function calls, RAG, admin UI, registration, metrics, and cron.
• Multimodal examples – screenshots in the README show image and voice interactions.

---

Who might use this?

• Developers who need a ready‑made bridge between popular chat apps and any LLM API.
• Teams deploying internal AI assistants across multiple corporate messengers.
• Hobbyists who want a single bot that can answer questions, generate images, or run scheduled LLM tasks.

---

License

MIT (as indicated by the badge).

---

Bottom line – MuseBot is a practical, Go‑based framework for turning everyday messaging platforms into AI chat agents, with support for many LLM back‑ends, streaming replies, multimodal input, and operational features like RAG, metrics, and cron scheduling.