Skip to main content

​
Building nervous system for agent context

GitNexus indexes any codebase into a knowledge graph — every dependency, call chain, cluster, and execution flow — then exposes it through smart tools so AI agents never miss code.

​
The problem

Tools like Cursor, Claude Code, Cline, Roo Code, and Windsurf are powerful — but they don’t truly know your codebase structure. What happens:
  1. AI edits UserService.validate()
  2. Doesn’t know 47 functions depend on its return type
  3. Breaking changes ship
GitNexus solves this by precomputing relational intelligence at index time — clustering, tracing, scoring — so tools return complete context in one call.

​
Two ways to use GitNexus

CLI + MCP

Index repos locally, connect AI agents via MCP. For daily development with Cursor, Claude Code, Windsurf, OpenCode.

Web UI

Visual graph explorer + AI chat in browser. No install required — drag & drop and start exploring.

​
Key features

Multi-repo MCP server

One MCP server serves all indexed repos. Set it up once, works everywhere.

Process-grouped search

Hybrid search (BM25 + semantic) returns results grouped by execution flows, not just file matches.

Impact analysis

Blast radius analysis shows what breaks at depth 1/2/3 with confidence scores.

360-degree context

See all callers, callees, imports, and process participation for any symbol in one view.

Git-diff impact

Maps changed lines to affected processes — know the risk before you commit.

Multi-file rename

Coordinated rename with graph + text search, confidence-tagged edits for review.

11 languages

TypeScript, JavaScript, Python, Java, C, C++, C#, Go, Rust, PHP, Swift.

100% local

Everything runs locally. No network calls. Your code never leaves your machine.

​
What your AI agent gets

When you connect GitNexus via MCP, your AI agent gets:
  • 7 tools: list_repos, query, context, impact, detect_changes, rename, cypher
  • 6 resource types: repos, clusters, processes, schema, and more
  • 2 MCP prompts: detect_impact, generate_map
  • 4 agent skills: exploring, debugging, impact analysis, refactoring
See the MCP integration guide for full details.

​
Get started

1

Index your repository

Run this from your repo root:
2

Configure MCP

Auto-detect and configure your editors:
3

Start coding

Your AI agent now has deep codebase awareness. Try asking it to trace a function or analyze impact.
Ready to dive in? Check out the quickstart guide.

​
Community

Discord

Join the official Discord to discuss ideas, issues, and more.

GitHub

Star the repo, report issues, or contribute.

​
Star history

Star History Chart

​
Privacy & security

  • CLI: Everything runs locally on your machine. No network calls. Index stored in .gitnexus/ (gitignored). Global registry at ~/.gitnexus/ stores only paths and metadata.
  • Web: Everything runs in your browser. No code uploaded to any server. API keys stored in localStorage only.
  • Open source — audit the code yourself.