thuongh2

thuongh2 / git-mimir

Public

Mimir is a Go-based code intelligence engine that indexes your repository into a knowledge graph and exposes it to AI agents via MCP, HTTP API, and an interactive web UI.

github.comthuongh2git-mimir ai graphql-database mcp-server
11
4
100% credibility
Found Mar 19, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Go
AI Summary

Mimir builds an interactive knowledge graph of your codebase to help humans visually explore structure and AI agents analyze impacts.

How It Works

1
🔍 Discover Mimir

You hear about Mimir, a helpful tool that turns confusing code into an easy-to-explore map of how everything connects.

2
📥 Get Mimir running

With one quick command, you download and set it up on your computer—no hassle, just ready to go.

3
📁 Point to your project

You simply tell Mimir the folder with your code, and it starts scanning everything inside.

4
✨ Magic map appears

In seconds, Mimir builds a complete picture of your code—functions, classes, and how they link together.

5
Choose your way to explore
🖥️
Visual explorer

Open a web page to zoom around the colorful map, click symbols to see connections.

🤖
AI companion

Link it to your coding AI so it understands your project perfectly during chats.

6
🔗 Search and understand

Ask questions like 'what calls this function?' or 'what breaks if I change that?' and get clear answers.

🎉 Code clicks instantly

Now your project feels simple and familiar—you spot issues fast and make changes confidently.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 11 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is git-mimir?

Git-mimir is a Go-based code intelligence engine that scans Git repos, builds a knowledge graph of symbols and relationships across eight languages, and exposes it to AI agents via MCP, an HTTP API, and an interactive web UI. Developers run `mimir analyze /path/to/repo` for full or incremental indexing, then query with tools like `query("auth flow")` for semantic search or `impact("UserService")` for blast radius analysis. Unlike Grafana Mimir for metrics or its helm charts, this focuses on code graphs for agents and humans.

Why is it gaining traction?

It crushes alternatives on speed—cold indexes in under 8 seconds for 1,000 files, incremental updates in 2 seconds—thanks to pure Go with no CGO, tiny 20MB binaries, and low RAM use. The daemon mode (`mimir daemon start`) keeps MCP servers running for editors like Cursor or Claude Code, with auto-setup hooks and skills. Hybrid BM25+vector search via API or UI delivers sub-80ms results, making it a lightweight drop-in for go-based graph queries.

Who should use this?

AI agent users in VS Code, Cursor, or Zed building on monorepos, needing quick context for refactors or debugging. Backend teams tracing execution flows or analyzing git diffs before commits. Devs tired of grep in large TS/Go/Python bases, wanting visual graphs and MCP tools like rename or cypher for safe changes.

Verdict

Promising early tool for local code intel with agents—install via one-liner curl script and try on a small repo. At 11 stars and 1.0% credibility, it's raw with solid docs but light tests; watch for stability before production teams.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.