Your AI finally understands your codebase.

A MCP turns your codebase into a knowledge graph, giving AI tools instant access to functions, call chains, and context.

npx code-graph-builder@latest --setup

"Your AI assistant is brilliant — but it's navigating your codebase blind. It doesn't know your private APIs, your internal call chains, or which function actually handles that edge case. Code Graph Builder gives it a map."

← Swipe left to discard card and explore next →

Indexed once,
AI ready forever.

Step 1: graph-build Waiting to start...
Step 2: api-doc-gen Waiting to start...
Step 3: embed-gen Waiting to start...
def process_packet(data): ...
📄 API Doc: process_packet
How does packet parsing work?
Agent (RAG)
It uses process_packet. Found via semantic search in graph context.

Turn error logs into investigation worksheets.

"A log error fires. You have a function name. You have no idea what triggered it."

[2026-04-05 14:22:01] INFO: Scanning workspace codebase...
[2026-04-05 14:22:02] ERROR in GraphUpdater.run() at line 89
[2026-04-05 14:22:02] WARN: Graph builder process terminated

Explore your code graph.

Navigate deep call trees with mathematical precision. Every node is a real function from the code_graph_builder repo.

Inspector

Select a function to see details.

Quick Start

Two steps to supercharge your AI assistant.

Step 1: Install & Init

npx code-graph-builder@latest --setup

Step 2: Configure MCP

Add to your Claude Code / Cursor / Windsurf settings:

{
  "mcpServers": {
    "code-graph-builder": {
      "command": "npx",
      "args": ["-y", "code-graph-builder@latest", "--server"]
    }
  }
}

Supported Languages

Python • C / C++ • JS / TS • Rust • Go • Java • C#