code-intelligence-tools

Indexing your codebase into a knowledge graph reduces token consumption by ~59% and speeds up AI agent response times by ~49%. This structured approach lets agents query codebase data directly instead of scanning raw files.

CodeGraph works with Claude Code, Codex CLI, OpenCode, Hermes Agent, Cursor, Gemini CLI, Qwen Code, and GitHub Copilot CLI. It indexes codebase data so these AI agents can rapidly understand code structure and call graphs.

Run `codegraph init -i` in your project directory to initialize codebase indexing. This builds structured indexes including knowledge graphs, AST maps, and call graphs that AI agents can query for fast code understanding.

A code knowledge graph is a structured index of codebase relationships, AST maps, and call graphs. It speeds up AI coding by allowing agents to query specific code paths directly, reducing the need to process entire files and cutting response times significantly.

Yes, you need the CodeGraph dependency installed to enable indexing and querying of codebase data. This Skill relies on CodeGraph to build the structured knowledge graphs that AI agents use to understand your code efficiently.

Codebase indexing with knowledge graphs requires an initialization step and depends on the CodeGraph tool. While it reduces token usage and speeds up responses, you must run the indexing process first and ensure compatibility with your specific AI agent.