Klear Team Brain

What it does

Klear Team Brain turns fragmented AI coding sessions (from Claude Code or Codex) into a shared, searchable team memory. It automatically collects session transcripts, mirrors team documentation from sources like Lark/Feishu or Notion, and integrates live GitHub code state into a single git-based truth store.

Tools

• grep: Search content across session transcripts and mirrored docs using regex.
• find: Locate specific files or sessions by name or glob patterns.
• read: Read the contents of a specific session or document by path.
• ls: Inspect the structure of the truth store, including available workspaces and branches.
• sessions: Filter and find AI sessions by person, repository, or time window.
• stats: View aggregate token usage and activity trends for the team.
• log: Browse the forensic activity timeline of the memory store.
• read_github: Fetch live code state or specific file contents directly from GitHub/GitLab.

Installation

Clone the repository and run the quickstart:

git clone https://github.com/Asklear/Klear-Team-Brain.git && cd Klear-Team-Brain && npm install && npm run quickstart

For MCP configuration (stdio):

{
  "mcpServers": {
    "team-brain": {
      "command": "node",
      "args": ["/path/to/Klear-Team-Brain/mcp/server.mjs"]
    }
  }
}

Supported hosts

• claude
• codex
• cursor
• gemini-cli