CodeGraph is a pre-indexed code knowledge graph that enhances AI coding agents with semantic code intelligence, providing a 100% local, fast, and surgical code context for developers.
Source: Official README and linked documentation View on GitHub →CodeGraph is being discussed because its integration with popular AI coding agents like Claude Code, Cursor, Codex, and GitHub Copilot, offering a unique local code intelligence solution that reduces tool calls and enhances developer productivity. Its Rust-powered kernel and auto-sync feature is unique technical choices.
Source: Official README and linked documentationCodeGraph indexes codebases to create a knowledge graph that provides semantic code intelligence, enabling faster and more accurate code analysis and context-aware suggestions.
Source: Official README and linked documentationThe graph automatically updates in real-time as code changes, ensuring the index is always current and relevant.
Source: Official README and linked documentationCodeGraph integrates with a variety of AI coding agents, enhancing their capabilities with semantic code intelligence.
Source: Official README and linked documentationCodeGraph operates locally, ensuring privacy and reducing latency associated with remote processing.
Source: Official README and linked documentationThe architecture of CodeGraph is inferred to be modular, with a Rust kernel for performance-critical tasks, TypeScript for the main application logic, and a CLI for user interaction. The code tree indicates a separation of concerns with distinct directories for different functionalities, such as `.claude/skills` for Claude Code integration and `__tests__` for testing.
Source: Official repository tree and dependency files@clack/promptscommanderfast-string-widthfast-wrap-ansiignorejsonc-parserpicomatchsisteransiCodeGraph is suitable for developers using AI coding agents in their workflow. It is useful in scenarios where developers need to enhance their coding productivity, understand complex codebases quickly, and maintain code quality.
Source: Official README and linked documentationThis review follows the current feature description in the official README and does not add a release version that was not verified.
Source: github.com/codegraph-ai/CodeGraphstatus: docs-only | checked_at: 2026-09-23 | environment: Reviewed the official README and its descriptions of MCP, VS Code, and supported language counts. | limitations: Language parsing, indexing time, graph accuracy, and MCP calls were not tested. | observed_output: The README claims 42 MCP tools, 38 languages, and a persistent memory layer.
Source: Official README and linked documentationCompared with grep, CodeGraph can answer questions about relationships across files. Compared with language servers, it places more emphasis on a persistent graph and agent queries. These tools are not mutually exclusive, and important conclusions still need confirmation through source code, tests, and builds.
Source: Official README and linked documentationCodeGraph is a project for developers looking to integrate semantic code intelligence into their workflow with AI coding agents. Its unique local execution and auto-sync capabilities make it a tool for enhancing productivity and understanding complex codebases.
The README claims 42 MCP tools, 38 languages, and a persistent memory layer. Language parsing, indexing time, graph accuracy, and MCP calls were not tested.
The review environment was: Reviewed the official README and its descriptions of MCP, VS Code, and supported language counts. codegraph behavior outside that scope still needs a local check.