hedwig-ai

hedwig-ai / hedwig-code-graph

Public

Build and Search a knowledge base for your projects—bringing together code, PDFs, Markdown, and everything else.

27
2
100% credibility
Found Apr 14, 2026 at 26 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

hedwig-cg builds a local searchable map of codebases and documents to help AI coding tools understand large projects through hybrid search.

How It Works

1
📰 Discover hedwig-cg

You hear about a helpful tool that creates a full map of your codebase so AI assistants can understand every part of your project.

2
💻 Add to your project

You simply add the tool to your coding folder with one easy step.

3
🤝 Connect your AI helper

Choose your favorite AI coding buddy like Claude or Cursor and link it up in seconds.

4
🗺️ Build the code map

Ask your AI to scan your project and create a smart map of all functions, connections, and notes.

5
🔍 Search and explore

Now search anything in your project, and your AI finds exact matches with context instantly.

6
🔄 Stays up to date

The map updates automatically whenever you change code, keeping everything fresh.

🎉 AI superpowers unlocked

Your AI now grasps your entire project perfectly, giving spot-on advice and saving you tons of time.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 26 to 27 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 hedwig-code-graph?

Hedwig-code-graph is a Python tool that builds a searchable knowledge base from your project's code, PDFs, Markdown, and configs, creating a graph of functions, classes, and docs. It solves the problem of AI coding agents only seeing file snippets by enabling hybrid search across your entire 10k+ file codebase—vector, graph, keyword, and community signals fused locally with SQLite and FAISS. Run `hedwig-cg build .` once, then query via CLI or integrate with agents.

Why is it gaining traction?

One-command installs hook into Claude Code, Cursor, Aider, and others, auto-rebuilding on changes and prompting agents to search the graph before grepping. Dual-model embeddings (code/text) plus git co-change edges give precise results without cloud APIs, and incremental builds skip unchanged files for 95% speedups. It's a local-first way to build search index for AI apps, beating naive keyword tools on structure-aware queries.

Who should use this?

Backend devs building GitHub Copilot agents or GitHub apps on monorepos, where agents need full context. Frontend teams maintaining GitHub Pages sites or portfolios with scattered docs/PDFs. Anyone tackling build search engine from scratch challenges, like Google’s search and recommendations AI labs, on local projects.

Verdict

Try it if you pair AI agents with large codebases—installs fast, runs offline, delivers god-node stats and hybrid search out of the box. At 20 stars and 1.0% credibility, it's alpha (solid docs, PyPI-ready, but light tests); build your GitHub project graph today, watch for polish.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.