MCP Catalogs
← Homechunkhound screenshot

chunkhound

by chunkhound·★ 1,258·Score 55

Local-first codebase intelligence tool with MCP integration for semantic code searching across 32+ languages.

developer-toolsai-llmsearch
99
Forks
82
Open issues
this month
Last commit
2d ago
Indexed

Overview

ChunkHound is a local-first codebase intelligence tool that extracts architecture, patterns, and institutional knowledge from codebases. It uses a research-backed cAST algorithm for semantic code chunking and provides multi-hop semantic search capabilities beyond direct matches. The tool supports both semantic search using natural language queries and regex pattern matching without API keys, making it versatile for different use cases while maintaining local privacy.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:Searching large monorepos with cross-team dependencies
you:Security-sensitive codebases requiring local-only processing
you:Multi-language projects needing consistent search capabilities
you:Does ChunkHound require internet connectivity?
you:Which programming languages does ChunkHound support?

When to choose this

Choose ChunkHound when you need deep code relationship analysis in security-sensitive environments or large multi-language codebases where traditional search falls short.

When NOT to choose this

Avoid ChunkHound if you need cloud-based search with minimal local setup or if you're working exclusively with codebases that don't benefit from semantic relationship analysis.

Tools this server exposes

6 tools extracted from the README
  • search_code

    Semantic search for code using natural language queries

  • regex_search

    Pattern matching search in code without API keys

  • index_codebase

    Index the codebase for semantic search and analysis

  • find_dependencies

    Discover interconnected code relationships and dependencies

  • code_analysis

    Analyze code patterns and architecture

  • research_code

    Deep research into code using multi-hop semantic search

Note: Tool names inferred from feature descriptions and documentation about capabilities. The README mentions MCP integration but doesn't explicitly list tool names with signatures.

Comparable tools

semantic-sensecodesearchast-grepripgrep

Installation

  1. Install uv package manager if needed: curl -LsSf https://astral.sh/uv/install.sh | sh
  2. Install ChunkHound: uv tool install chunkhound
  3. Create .chunkhound.json in project root with your API keys (optional for regex search)
  4. Index your codebase: chunkhound index

For Claude Desktop integration, add to your config.json:

{
  "mcpServers": {
    "chunkhound": {
      "command": "chunkhound",
      "args": ["serve"]
    }
  }
}

FAQ

Does ChunkHound require internet connectivity?
No, ChunkHound is local-first and works offline. However, some features like semantic search may require API keys for embeddings which would need internet.
Which programming languages does ChunkHound support?
ChunkHound supports 32 languages including Python, JavaScript/TypeScript, Java, C/C++, Go, Rust, and many others. It uses Tree-sitter for parsing programming languages and custom parsers for text-based files.

On Hacker News

Recent discussion from the developer community.

Compare chunkhound with

chunkhound vs ultimate_mcp_serverchunkhound vs mcp-server-chartchunkhound vs everythingchunkhound vs filesystemchunkhound vs time
GitHub →

Last updated · Auto-generated from public README + GitHub signals.