# chunkhound/chunkhound

Local first codebase intelligence

Repository: https://github.com/chunkhound/chunkhound
Canonical: https://ross.abutalabs.com/products/chunkhound
Homepage: https://chunkhound.ai
Language: Python
License: MIT
License Family: permissive
Topics: agent, ai, duckdb, mcp-server, semantic-search, tree-sitter, rag
Last push: 2026-08-21T11:42:46+00:00

## Health v2 (maintenance only)
Score: 83/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 98, release rhythm 92, longevity 32
- inputs: {"age_days": 455, "days_push": 12, "days_rel": 52, "gap_med": 0, "n_releases_24m": 42}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1418, forks 129 (observed 2026-08-28T04:04:40.151960+00:00)

## What it is
ChunkHound is a local-first codebase intelligence CLI that indexes code with tree-sitter parsing and semantic/regex search, exposing cited context to AI agents via MCP. It combines current code, git history, and web research into implementation-backed answers for agents and teams.

## Use cases
- give coding agents cited context about how a codebase works before editing
- search my codebase semantically and with regex locally
- summarize what changed in a branch or release for code review
- trace a stack trace or bug report to the relevant code paths
- run an MCP server that lets Claude Code or Cursor query my repo
- index a monorepo for AI deep research across many languages

## When to choose
- you want local-first code search and RAG without your code leaving the machine
- you use MCP-compatible agents like Claude Code, Cursor, or Codex and need deep codebase context
- you need cited answers combining code, git history, and external docs

## When to avoid
- you need a hosted or team-shared code search service rather than a local CLI
- you only need simple grep-style search without semantic indexing
- you cannot install Python 3.10+ and uv

## Facets
- artifact type: cli-tool
- maturity: active
- function: search-engine, rag, mcp, parser, llm-inference, developer-tools
- domain: developer-tools, artificial-intelligence, version-control
- platform: cli, python, cross-platform, windows
- tags: code-search, semantic-search, tree-sitter, duckdb, codebase-intelligence, local-first, git-history, mcp-server, code-chunking, ai-agents, retrieval-augmented-generation, search, macos, linux

## Member repositories
- chunkhound/chunkhound (main) score 83

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:40.151960+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:37:56.445709+00:00, confidence not recorded.
  - readme: https://github.com/chunkhound/chunkhound (fetched 2026-08-28T04:04:40.151960+00:00, sha f5673d3fcb8d)
  - homepage: https://chunkhound.ai (fetched 2026-08-29T11:50:25.746485+00:00, sha 3cc2cecd150f)
  - site_page: https://chunkhound.ai/docs/getting-started (fetched 2026-08-29T11:50:25.755715+00:00, sha 2e87618e40d5)
  - site_page: https://chunkhound.ai/docs/configuration (fetched 2026-08-29T11:50:25.757651+00:00, sha aa26bd7f37cc)
  - registry_pypi: https://pypi.org/pypi/chunkhound/json (fetched 2026-08-29T11:50:25.760381+00:00, sha 209cd58c628f)
- Data as of 2026-08-30T08:39:29.467469+00:00.