MCP Catalogs
Home

ClaudeHistoryMCP

by jhammant·65·Score 48

MCP server for searching Claude Code conversation history with hybrid BM25 + TF-IDF search and knowledge extraction.

ai-llmdeveloper-toolssearch
13
Forks
1
Open issues
3 mo ago
Last commit
2d ago
Indexed

Overview

ClaudeHistoryMCP is a specialized MCP server that indexes Claude Code conversation history to make it searchable and reusable. It uses a hybrid search approach combining BM25 for keyword precision and TF-IDF for semantic recall, with Reciprocal Rank Fusion to combine results. The server automatically extracts knowledge from conversations, including decisions, solutions, and error-fix patterns, and can inject project context at the start of new sessions. It supports filter syntax for targeted searches and includes cloud sync capabilities for knowledge sharing.

Try asking AI

After installing, here are 6 things you can ask your AI assistant:

you:Debugging recurring issues by searching past solutions to similar problems
you:Retrieving project decisions and context when starting new sessions
you:Discovering patterns in development workflows across multiple projects
you:Does this work with all Claude Code conversations?
you:How large can the history database be?
you:Can I sync knowledge across devices?

When to choose this

Choose ClaudeHistoryMCP when you work frequently with Claude Code and need to efficiently search and retrieve past solutions, decisions, and patterns from your conversation history.

When NOT to choose this

Avoid this if you're not using Claude Code, as it's specifically designed to work with Claude's conversation format and stores data in Claude's directory structure.

Tools this server exposes

9 tools extracted from the README
  • search_history

    Full-text search across all conversations with filter syntax

  • find_solutions

    Find how you fixed errors/problems before

  • get_session_summary

    Structured summary of any session

  • list_projects

    List all projects with session counts and dates

  • find_patterns

    Discover recurring topics, workflows, and issues

  • get_project_context

    Full project context (recent sessions, decisions, knowledge)

  • cloud_sync_push

    Push knowledge and sessions to cloud server

  • cloud_sync_pull

    Pull knowledge and sessions from cloud server

  • cloud_sync_status

    Check cloud sync configuration and connection

Comparable tools

claude-desktop-mcpconversation-search-mcpclaude-code-history-search

Installation

Installation

  1. Clone and build the server:
git clone https://github.com/jhammant/ClaudeHistoryMCP.git
cd ClaudeHistoryMCP
npm install
npm run build
  1. Build the search index:
npm run build-index
  1. Register the MCP server with Claude:
claude mcp add claude-history -- node "/path/to/ClaudeHistoryMCP/dist/index.js"
  1. Install the session-start hook (optional):
npm run install-hook
  1. Add to your global CLAUDE.md (recommended):
## Claude History MCP

When the `claude-history` MCP is available, use it proactively:
- **Session start**: Use `get_project_context` to check for prior decisions, patterns, and recent session summaries
- **Debugging**: Use `find_solutions` to search history for past fixes before starting from scratch

FAQ

Does this work with all Claude Code conversations?
Yes, it indexes all JSONL conversation files stored in ~/.claude/projects/
How large can the history database be?
Tests show ~9s index build time for ~170 sessions, with searches <200ms
Can I sync knowledge across devices?
Yes, using the optional cloud sync feature with ClaudeHistory Cloud

Compare ClaudeHistoryMCP with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.