Beever-AI/beever-atlas

What it does

Beever-Atlas is an open-source MCP server that converts conversations from Slack, Discord, Teams, or Mattermost into a structured knowledge graph and auto-generated wiki. It enables cited answers, semantic search, expert finding, and decision tracing across your team's chat history.

How to set up

Installation involves: 1) Cloning the BeeverAI/beeveratlas repository from GitHub, 2) Installing dependencies using Python package management, 3) Configuring your chat platform credentials (Slack, Discord, Teams, or Mattermost), 4) Setting up your LLM provider via LiteLLM (bringing your own model), 5) Running the MCP server to begin indexing chat data.

Key features

• Automatic conversion of chat conversations into typed knowledge graphs
• Auto-generated wiki with properly cited answers
• Semantic search capabilities across all indexed content
• Expert finding to identify team members with specific knowledge
• Decision tracing to track reasoning and conclusions
• Flexible LLM support through LiteLLM integration (BYOM)
• Multi-platform chat integration (Slack, Discord, Teams, Mattermost)
• Open-source under Apache 2.0 license