npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

git-memory

v0.2.2

Published

Semantic Git history index for Claude Code — Node.js port of claude-memory

Readme

git-memory

A semantic index over your Git commit history for Claude Code. Gives Claude an MCP server with 5 tools to query what changed, why, and where — before it edits your code.

Node.js port of claude-memory (Python).


How it works

git-memory walks your Git log, filters signal-rich commits, embeds them with a local or cloud embedding model, and stores them in ChromaDB. An MCP server exposes 5 query tools to Claude Code so it can answer "what's the history of this module?" before touching it.

git history → filter → embed → ChromaDB → MCP tools → Claude Code

Requirements

  • Node.js 18+
  • Git repository to index
  • One of the following embedding providers:
    • Ollama (default) — run ollama pull nomic-embed-text and ollama serve
    • OpenAI — set OPENAI_API_KEY
    • fastembed — fully local, no server needed

Installation

npm install -g git-memory

Or use directly without installing:

npx git-memory index --repo-path /path/to/repo --user-id my-repo

Quick Start

1. Index your repository

# Dry run first — see what will be indexed
git-memory index --repo-path /path/to/repo --user-id my-repo --dry-run

# Full index
git-memory index --repo-path /path/to/repo --user-id my-repo

From within the repo directory:

cd /path/to/repo
git-memory index --user-id my-repo

For large repos (>1000 commits), limit to recent history:

git-memory index --user-id my-repo --limit 500

2. Install the MCP server into Claude Code

git-memory install --repo-path /path/to/repo --user-id my-repo

This writes the MCP server config to .claude.json in your repo.

3. Restart Claude Code

The MCP tools are now available in Claude Code sessions for that repo.


MCP Tools

Once installed, Claude Code has access to these 5 tools:

| Tool | Description | |------|-------------| | search_git_history(query, limit, category) | Semantic search over commit messages and metadata | | latest_commits(limit) | Most recent indexed commits | | commits_touching_file(filename, limit) | All commits that modified a file | | bug_fix_history(component, include_security) | Bug and security fixes for a component | | architecture_decisions(topic, limit) | Refactors, migrations, design decisions |

Category filter

Pass category= to search_git_history to narrow results:

| Value | Matches | |-------|---------| | fix | Bug fixes, hotfixes, patches | | feat | New features | | security | Security-related changes | | refactor | Code refactors | | migration | Database/schema migrations | | arch | Architecture decisions | | perf | Performance improvements |


Check status

git-memory status --repo-path /path/to/repo

Expected output:

── git-memory status ──────────────────────────
  Repo          : my-repo
  Chroma docs   : 39 (this repo) / 39 (all repos)
  Total commits : 42
  Coverage      : 93%
  Embed provider: ollama

Configuration

All options can be set via environment variables or CLI flags.

| Environment Variable | CLI Flag | Default | Description | |---------------------|----------|---------|-------------| | GIT_MEMORY_REPO_PATH | --repo-path | cwd | Path to the Git repo | | GIT_MEMORY_USER_ID | --user-id | git_memory_system | Unique ID per repo (used to namespace the index) | | GIT_MEMORY_EMBED_PROVIDER | — | ollama | Embedding provider: ollama, openai, fastembed | | GIT_MEMORY_EMBED_MODEL | — | nomic-embed-text | Embedding model name | | GIT_MEMORY_LLM_MODEL | — | qwen2.5 | LLM for context enrichment (Layer 2) | | OLLAMA_URL | — | http://localhost:11434 | Ollama server URL | | OPENAI_API_KEY | — | — | Required when using OpenAI provider | | GIT_MEMORY_CHROMA_DIR | — | ~/.cache/git_memory/chroma_commits | ChromaDB storage path | | GIT_MEMORY_CONTEXT_DIR | — | ~/.cache/git_memory/chroma_context | Context store path |

Embedding providers

Ollama (default)

ollama pull nomic-embed-text
ollama serve
git-memory index --user-id my-repo

OpenAI

export OPENAI_API_KEY=sk-...
export GIT_MEMORY_EMBED_PROVIDER=openai
export GIT_MEMORY_EMBED_MODEL=text-embedding-3-small
git-memory index --user-id my-repo

fastembed (fully local, no server)

export GIT_MEMORY_EMBED_PROVIDER=fastembed
# Uses BAAI/bge-small-en-v1.5 by default
git-memory index --user-id my-repo

MCP config reference

The install command writes this to .claude.json automatically. You can also configure it manually:

{
  "mcpServers": {
    "git-memory": {
      "command": "git-memory",
      "args": ["serve"],
      "env": {
        "GIT_MEMORY_REPO_PATH": "/path/to/repo",
        "GIT_MEMORY_USER_ID": "my-repo-name",
        "GIT_MEMORY_EMBED_PROVIDER": "ollama",
        "GIT_MEMORY_EMBED_MODEL": "nomic-embed-text",
        "GIT_MEMORY_LLM_MODEL": "qwen2.5",
        "OLLAMA_URL": "http://localhost:11434"
      }
    }
  }
}

Claude Code skills

git-memory ships four Claude Code skills in the skills/ directory. These guide Claude on when and how to use the MCP tools:

| Skill | Trigger phrase | |-------|---------------| | git-memory-index | "index this repo", "set up git memory" | | git-memory-search | "why was this written this way?", "find commits touching auth" | | git-memory-debug | "why did this break?", "trace all changes to OrderService" | | git-memory-status | "is git memory set up?", "how many commits are indexed?" |

Add this to your CLAUDE.md to activate them:

When beginning any task:
1. Call `latest_commits(5)` to understand recent changes
2. Call `search_git_history(<topic>)` before editing any module
3. After fixing a bug, call `bug_fix_history(<component>)` to check for prior regressions

Development

git clone https://github.com/YOUR_ORG/git-memory
cd git-memory
npm install

# Run CLI without building
npm run dev -- index --user-id test

# Start MCP server in dev mode
npm run serve:dev

# Build TypeScript
npm run build

Troubleshooting

| Symptom | Fix | |---------|-----| | 0 commits indexed | Run git-memory index --repo-path . | | MCP tools return errors | Check Ollama is running: ollama serve | | Wrong user-id | Must match GIT_MEMORY_USER_ID in MCP config | | Search returns nothing | Confirm index has data: git-memory status | | Layer 2 context missing | Normal if LLM is unavailable — Layer 1 (ChromaDB) still works |


License

AGPL-3.0-or-later