ChunkHound
Local code search and research with citations: semantic search, cross-file behavior explanation and git-history queries
Medium risk
We rate an entry medium when the tool runs code, makes network calls or reads project files. Check what exactly it does before installing.
Why this level
- Reads and indexes code files
- May send code to an external provider for semantic search
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/chunkhoundDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Install the tool
pipx install chunkhoundWithout pipx, pip install chunkhound works too.
Install: uv tool install chunkhound, then chunkhound index . and chunkhound research "question". For semantic search set an embedding provider in .chunkhound.json.
Other ways from the author
uv tool install chunkhoundAfter installing, run chunkhound index . in the project directory.
This is third-party code. Review the repository files before installing.
What it does
ChunkHound indexes a codebase and lets you search it by meaning, not only by text. It explains behavior across several files with source citations and can answer questions about git history: by last commits, tag, branch or range. Indexing is local and stays under your control. Regex search works without external services, while semantic search and deep research need an embedding provider and a model.
Who it is for. For developers who need semantic search and analysis of a large or unfamiliar codebase.
Good fit when
- You need to find code by meaning rather than exact text
- You need behavior explained with file citations
- You need to understand a large PR from git history
Not a fit when
- Plain text search over the project is enough
- You cannot send code to an external embedding provider
Example request
Index the project with ChunkHound and explain, with citations, how authentication worksLimitations
Semantic search and deep research need an embedding provider and a model, some of them paid, while regex search works without them. The index has to be built, which takes time on a large codebase. You can choose local providers to keep code from leaving the machine.
How to disable. Remove the tool with uv tool uninstall chunkhound and delete the created index.
Security check
- Reads and indexes code files
- May send code to an external provider for semantic search
README in short
The README describes ChunkHound as a local code search and research tool with source citations. Installation is via uv, and the index and research commands build an index and give grounded answers. Regex search works without a provider, while semantic search and deep research need an embedding provider and a model set in .chunkhound.json. Many languages and git-history search are supported.
FAQ
Can it run without external services?
Regex search works locally without a provider. Semantic search and research need an embedding provider and a model, and you can pick local ones.
Does git-history search work?
Yes, you can ask by last commits, hash, tag, branch or range.
Related
A skills library that gives coding agents a development process: brainstorming, planning, TDD, subagents and code review
Skills for real engineers by Matt Pocock
Skills For Real Engineers
Small composable skills for engineering with agents: plan grilling, TDD, bug diagnosis, code review and architecture
GitHub toolkit for spec-driven development: the specify CLI adds agent commands and skills to a project, from principles to implementation
Reference MCP servers
Model Context Protocol servers
Official reference MCP servers: Filesystem, Fetch, Git, Memory, Sequential Thinking, Time and Everything