Codanna

v1.0.0Developer Toolsstable

Local code intelligence MCP server and CLI for AI coding agents

ai-agentscall-graph-analysiscode-indexingcode-intelligencecode-search
Share:
675
Stars
0
Downloads
0
Weekly
0/5

What is Codanna?

Codanna is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to local code intelligence mcp server and cli for ai coding agents

Local code intelligence MCP server and CLI for AI coding agents

This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Local code intelligence MCP server and CLI for AI coding age

Use Cases

Index local code repositories for AI agents with fast, local-first search. Perform call graph analysis and code relationship detection. Search code by meaning and structure without cloud dependencies.
bartolli

Maintainer

LicenseApache-2.0
Languagerust
Versionv1.0.0
UpdatedMay 21, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx codanna

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Codanna

Codanna is a local-first, Rust-powered code intelligence MCP server that indexes your codebase and exposes semantic search, call graph analysis, and dependency tracking to AI coding agents. It parses 13+ programming languages at 75,000+ symbols per second using Tree-sitter, stores the index locally for sub-10ms lookup times, and requires no cloud services or API keys. Developers use it with Claude Code, Cursor, and Windsurf to give AI agents accurate, real-time understanding of large codebases for navigation, refactoring, and documentation queries.

Prerequisites

  • Codanna binary installed (via install script, Homebrew, or Nix — no Rust toolchain needed for binary installs)
  • Approximately 150MB disk space for the local embedding model (auto-downloaded on first run)
  • An MCP client such as Claude Code, Cursor, or Windsurf
  • Linux users: pkg-config and libssl-dev system packages installed
  • No API keys or cloud accounts required
1

Install Codanna

Use the recommended install method for your platform. The curl installer handles binary download and PATH setup automatically.

# macOS / Linux / WSL:
curl -fsSL --proto '=https' --tlsv1.2 https://install.codanna.sh | sh

# Homebrew (macOS):
brew install codanna

# Windows (PowerShell):
irm https://raw.githubusercontent.com/bartolli/codanna/main/scripts/install.ps1 | iex

# Nix:
nix run github:bartolli/codanna
2

Initialize and index your project

Run codanna init in your project root to set up the local index, then index your source directory. The embedding model is downloaded automatically on first run (~150MB).

cd /path/to/your/project
codanna init
codanna index src
3

Configure Codanna as an MCP server in Claude Code

Add Codanna to your MCP client configuration. For Claude Code, add it to your project's .mcp.json or global MCP config.

{
  "mcpServers": {
    "codanna": {
      "command": "codanna",
      "args": ["mcp"]
    }
  }
}
4

Index documentation for RAG queries (optional)

Add and index documentation collections to enable document-level search alongside code search.

codanna documents add-collection docs ./docs
codanna documents index
5

Verify the MCP server is working

Test the MCP tools from the command line to confirm indexing and search are working correctly.

codanna mcp semantic_search_with_context query:"where do we handle errors" limit:3
codanna mcp search_documents query:"authentication flow"

Codanna Examples

Client configuration

Add this configuration to your Claude Code .mcp.json or Cursor MCP settings to enable Codanna code intelligence.

{
  "mcpServers": {
    "codanna": {
      "command": "codanna",
      "args": ["mcp"]
    }
  }
}

Prompts to try

Once Codanna is indexed and connected, use these prompts to leverage code intelligence:

- "Find all functions in this codebase that handle authentication"
- "Where is the database connection initialized and what calls it?"
- "Show me the call graph for the checkout() function"
- "Which files would be affected if I change the UserRepository interface?"
- "Find all places where HTTP errors are caught and explain the patterns"

Troubleshooting Codanna

Install script fails on Linux with SSL or pkg-config errors

Install the required system packages first: 'sudo apt-get install pkg-config libssl-dev' on Debian/Ubuntu, or 'sudo yum install pkgconfig openssl-devel' on RHEL/CentOS. Then re-run the install script.

Embedding model download fails or is slow

Codanna auto-downloads a ~150MB embedding model on first run. If the download fails, check your internet connection and disk space. The model is cached locally after the first download and not re-downloaded on subsequent runs.

Semantic search returns poor results after adding new code

Re-run 'codanna index src' to update the index after significant code changes. The index is not automatically updated when files change unless you have configured a file watcher.

Frequently Asked Questions about Codanna

What is Codanna?

Codanna is a Model Context Protocol (MCP) server that local code intelligence mcp server and cli for ai coding agents It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Codanna?

Follow the installation instructions on the Codanna GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Codanna?

Codanna works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Codanna free to use?

Yes, Codanna is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.

Browse More Developer Tools MCP Servers

Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "codanna": { "command": "npx", "args": ["-y", "codanna"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Codanna?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides