MCP Integration Overview

Connect GitNexus to your AI editor using the Model Context Protocol

GitNexus exposes its knowledge graph through the Model Context Protocol (MCP), enabling AI editors to query code intelligence directly during conversations.

What is MCP?

The Model Context Protocol is a standard interface for connecting AI assistants to external data sources and tools. GitNexus implements MCP to provide:

• 7 code intelligence tools for querying the knowledge graph
• 7 resources for structured on-demand data
• 2 prompts for guided workflows (impact analysis, architecture mapping)

Available Tools

Your AI agent gets these tools automatically when connected:

Tool	Purpose	Multi-Repo
list_repos	Discover all indexed repositories	—
query	Process-grouped hybrid search (BM25 + semantic + RRF)	Optional
context	360-degree symbol view with categorized refs and process participation	Optional
impact	Blast radius analysis with depth grouping and confidence scores	Optional
detect_changes	Git-diff impact — maps changed lines to affected processes	Optional
rename	Multi-file coordinated rename with graph + text search	Optional
cypher	Raw Cypher graph queries against the KuzuDB database	Optional

Multi-Repo Support: With one indexed repo, the repo parameter is optional. With multiple repos, specify which one: query({query: "auth", repo: "my-app"}).

Available Resources

Lightweight reads (100-500 tokens) for navigation:

Resource	Content
gitnexus://repos	List all indexed repositories (read first)
gitnexus://repo/{name}/context	Codebase stats, staleness check, available tools
gitnexus://repo/{name}/clusters	All functional clusters with cohesion scores
gitnexus://repo/{name}/cluster/{name}	Cluster members and details
gitnexus://repo/{name}/processes	All execution flows
gitnexus://repo/{name}/process/{name}	Full process trace with steps
gitnexus://repo/{name}/schema	Graph schema for Cypher queries

MCP Prompts

GitNexus provides two guided workflows:

Prompt	Purpose
detect_impact	Pre-commit change analysis — scope, affected processes, risk level
generate_map	Architecture documentation from the knowledge graph with mermaid diagrams

Editor Support

Editor	MCP	Skills	Hooks (auto-augment)	Support Level
Claude Code	✓	✓	✓ (PreToolUse)	Full
Cursor	✓	✓	—	MCP + Skills
Windsurf	✓	—	—	MCP
OpenCode	✓	✓	—	MCP + Skills

Claude Code gets the deepest integration: MCP tools + agent skills + PreToolUse hooks that automatically enrich grep/glob/bash calls with knowledge graph context.

Setup Options

Automatic Setup

Run gitnexus setup to auto-configure all detected editors

Claude Code

Full integration with MCP, skills, and PreToolUse hooks

Cursor

MCP and skills configuration for Cursor

Multi-Repo Usage

Global registry and multi-repo architecture

How It Works

When you run gitnexus analyze in a repository:

1. Indexes the codebase into a KuzuDB graph database (stored in .gitnexus/)
2. Registers the repo in ~/.gitnexus/registry.json (global registry)
3. Installs agent skills to teach AI agents how to use the tools
4. Creates context files (AGENTS.md, CLAUDE.md) for immediate use

The MCP server runs via npx gitnexus mcp and serves all indexed repositories from the global registry. Your editor spawns this process automatically when MCP is configured.

Next Steps

1. Index your repository

   Run npx gitnexus analyze in your repo root

2. Configure MCP

   Choose your setup method:

   • Run npx gitnexus setup for automatic configuration
   • Or follow the editor-specific guides
3. Start using tools

   Your AI agent now has access to all GitNexus tools and resources