Local code intelligence MCP server and CLI for AI coding agents, providing semantic search and call graph analysis.
codanna is easy to set up with strong trust signals. Check agent compatibility and use-case fit before adding it to your workflow.
gh repo view bartolli/codanna --webOpen the official README and confirm the supported install method.
Add the server entry to your MCP client config.
Restart your agent and verify that the server tools appear.
Codanna gives your AI coding assistant a deep understanding of your codebase. It can find functions by what they do, trace how code connects across files, and answer questions about your project instantly. It runs locally on your machine, so your code never leaves your computer.
Codanna is a local-first code intelligence tool that acts as an MCP (Model Context Protocol) server and CLI for AI coding agents. It provides structured knowledge of your codebase to AI assistants like Claude, Gemini, and Codex, enabling them to understand functions, trace relationships, and find implementations with surgical precision. Key features include semantic search (natural language queries against code and documentation), relationship tracking (call graphs, implementations, dependencies), document search (index markdown and text files for RAG workflows), and native MCP integration. It supports multiple languages including Rust, Python, JavaScript, TypeScript, Java, Kotlin, Go, PHP, C, C++, C#, Clojure, Lua, Swift, and GDScript. Performance is sub-10ms lookups with 75,000+ symbols/second parsing. Codanna is designed for rapid R&D and pair programming, providing instant answers when LSP is too slow. It requires approximately 150MB for the embedding model and runs on macOS, Linux, WSL, and Windows.
Strong trust signals; still review the README and permissions before production use.
Last commit was about 18 days ago.
687 GitHub stars indicate community interest.
7 open issues signal maintenance load.
Apache-2.0 license detected.
Find where a specific function is called across the entire codebase instantly.
Ask your AI assistant to locate authentication logic using natural language queries.
Analyze what breaks if you change a particular function by tracing its dependencies.
Index project documentation and query it alongside code for comprehensive context.
Integrate with Claude Code, Cursor, or Windsurf to enhance code understanding during pair programming.
Local code indexing may store source-derived metadata; keep indexes out of public artifacts.
Embedding or semantic search setup should be reviewed for local data retention and model/provider boundaries.
687
Stars
61
Forks
7
Issues
Apache-2.0
License
A project template that turns your repository into stable infrastructure for AI-assisted development.
MCP server that gives AI coding assistants persistent memory, semantic code search, and live telemetry.
Up-to-date code documentation for LLMs and AI code editors, eliminating outdated or hallucinated API references.
2 security/trust notes recorded.
Setup difficulty is 2/5.