claudememory

Give Claude Code a long-term memory of your Git history.

The problem

Claude Code is great at reading your current codebase — but it has no memory of why things were built the way they are.

Every session starts fresh. Ask it why a module was refactored, which bugs have hit the payment service before, or what architectural decision introduced the current state machine — and it has no answer. That context lives in your Git history, but it's inaccessible to the AI.

claudememory solves this. It indexes your entire Git commit history into a dual-layer semantic store and exposes it to Claude Code via MCP. Claude can now answer:

• "Has the auth module had bugs before?"
• "Why was this abstraction introduced?"
• "What changed in PaymentService over the last year?"
• "What commits touched this file?"

How it works

Two storage layers work together on every indexed commit:

ChromaDB acts as the ground truth (fast, accurate, deduplicated). Mem0 enriches results with learned context over time. Claude sees both merged together.

Install

pip install claudememory

No external services required. Ollama is optional — if it's not running, claudememory falls back to ChromaDB-only mode automatically.

Quick start

cd /your/repo

# 1. Index your repository (run from inside the repo)
claude-memory index

# 2. Install Claude Code skills + MCP config
claude-memory install

# 3. Restart Claude Code — then use /claude-memory-search, /claude-memory-debug etc.

After setup, Claude Code automatically has access to your commit history. No prompting required — it calls the tools proactively based on what you're working on.

MCP tools exposed to Claude

Claude Code skills

CLI reference

claude-memory index                    # bulk index all commits (run from repo root)
claude-memory install                  # install Claude Code plugin (run from repo root)
claude-memory status                   # show indexed coverage
claude-memory serve                    # start MCP server (stdio)
claude-memory store HEAD               # store single commit (post-commit hook)

# Optional flags
claude-memory index --repo-path /other/repo --user-id myapp
claude-memory status --repo-path /other/repo

Configuration

All settings via environment variables:

Better with conventional commits

claudememory automatically categorizes every commit by scanning its message for keywords. If you already use Conventional Commits (fix:, feat:, refactor:, etc.), the tool works out of the box with near-perfect accuracy.

Here's why it matters:

# Without conventional commits
"updated payment stuff"         → category: general (no signal)
"fixed the thing"               → category: fix ✓

# With conventional commits
"fix(payments): race condition" → category: fix ✓
"feat(auth): add OAuth2 flow"  → category: feat ✓
"refactor: extract BaseService" → category: refactor ✓
"perf(queries): add index"     → category: perf ✓

The categorizer maps directly to these prefixes:

Commits without any recognized keyword are still indexed but categorized as general. The relevance filter also uses these keywords to decide which commits are worth indexing at all, so well-written conventional commit messages mean higher coverage and better search results.

You don't need conventional commits to use claudememory. But if you adopt them, every tool gets smarter automatically.

Works great alongside GitNexus

These two tools answer different questions and are better together: