CocoIndex Code plugin
CocoIndex Code (ccc)
A plugin for AST-based semantic code search: a skill, an MCP server and the ccc CLI help the agent find code without reading whole files
- 2,980 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- Indexes the codebase contents and stores them in a local database
- May send code snippets to a cloud embedding provider if the local variant is not selected
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/cocoindex-code-cccDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Indexes the codebase contents and stores them in a local database
- May send code snippets to a cloud embedding provider if the local variant is not selected
This is third-party code. Review the repository files before installing.
What it does
CocoIndex Code indexes a codebase by its syntax trees and lets you search by meaning instead of exact text matches like plain grep. The agent gets a built-in ccc skill that initializes the index itself, keeps it updated after changes, and searches on request, saving tokens that would otherwise go into reading extra files. A separate ccc grep command does structural search by syntax pattern without building an index. The same engine can also run as an MCP server or be called straight from the terminal.
Who it is for. For developers who want their coding agent to find the right code in a large repository faster and with fewer tokens spent.
Good fit when
- The repository is large and plain grep cannot find code by the meaning of a query
- You want the agent to keep the code index up to date on its own
- You need structural search by code pattern without building an index
- You want one search tool shared across Claude Code, Codex, Cursor and Grok
Not a fit when
- The repository is small and plain text search already works fine
- You cannot spare about 1 GB for local embeddings or a paid key for a cloud provider
Example request
Find where the project handles database reconnection using semantic search, not an exact text matchLimitations
Local embeddings without a key require the heavier package variant with torch and transformers, around 1 GB. The slim variant only works with a cloud embedding provider and needs its API key. The index is stored in LMDB with a default 4 GiB cap that must be raised manually for very large codebases. It sends anonymous usage telemetry, which can be disabled with an environment variable.
How to disable. Remove the plugin via /plugin remove cocoindex-code (Claude Code), or uninstall the skill with npx skills remove cocoindex-code; delete the project's .cocoindex_code folder for a full cleanup.
FAQ
Do I need an API key?
Not if you install the [full] variant — it uses the local Snowflake/snowflake-arctic-embed-xs embedding model. The slim variant without [full] uses a cloud provider and needs its key.
How does ccc grep differ from ccc search?
ccc search searches by meaning through a vector index, while ccc grep searches by syntax-tree structure with no index or embeddings needed, similar to grep but for code shape rather than text.
Discussion
No questions or reviews yet.
Related
PluginsLocal code index for agents: the MCP tool map searches by meaning and structure, and code never leaves your machine
Source search ranks matches using code structure and gives agents relevant snippets without a separate search index.
The MCP server finds definitions and meaningful code fragments through a local index with vector and full-text search.
A local codebase index for agents: hybrid search, dependency graphs and change impact analysis
An MCP server that indexes a local repository with tree-sitter and gives the agent code search, file lookup and a structure summary of a file
A local MCP server that indexes your code so the agent pulls the functions it needs instead of rereading whole files
README in short
The README presents ccc as lightweight AST-based code search on top of the Rust-based CocoIndex engine, claiming roughly 70% token savings when searching. The repository doubles as a plugin marketplace for Claude Code and Grok, and contains the ccc skill, optional auto-indexing hooks, and an MCP server. It supports over 30 programming languages, installs via pipx, uv or a Docker image, and can be wired manually into Codex, OpenCode and Kilo Code as an MCP. Apache-2.0 licensed.
README badge
Original SKILL.md
--- name: ccc description: "This skill should be used when code search is needed (whether explicitly requested or as part of completing a task), when indexing the codebase after changes, or when the user asks about ccc, cocoindex-code, or the codebase index. Trigger phrases include 'search the codebase', 'find code related to', 'update the index', 'ccc', 'cocoindex-code'." --- # ccc - Semantic Code Search & Indexing `ccc` is the CLI for CocoIndex Code, providing semantic search over the current codebase and index management. ## Ownership The agent owns the `ccc` lifecycle for the current project — initialization, indexing, and searching. Do not ask the user to perform these steps; handle them automatically. - **Initialization**: If `ccc search` or `ccc index` fails with an initialization error, run `ccc init` from the project root directory, then `ccc index` to build the index, then retry the original command. - **Index freshness**: Keep the index up to date by running `ccc index` when the index may be stale. - **Installation**: If `ccc` itself is not found, refer to references/management.md for installation instructions.