CodeAlive MCP

The CodeAlive MCP server: the agent searches large codebases by meaning and exactly and follows call relations without a full read

  • 90 ★ on the repository
  • Last commit
MCP serverOfficial

Extended access

Why this level

  • The code of your repositories is indexed in the external CodeAlive service and read through the remote server
  • The API key opens access to the indexed repositories of the account
All reasons and checks

codealive-ai/codealive-mcp

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add mcp/codealive-mcp

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The code of your repositories is indexed in the external CodeAlive service and read through the remote server
  • The API key opens access to the indexed repositories of the account

This is third-party code. Review the repository files before installing.

What it does

CodeAlive is a context engine for large codebases built on graph retrieval, and this server exposes it to agents through MCP. The main tools: `semantic_search` searches by meaning, `grep_search` exactly by text, regular expressions and file names, `read_file` reads a file with a line range, `get_file_tree` shows the tree and `fetch_artifacts` loads the full code of found hits. Call, inheritance and reference relations are expanded by `get_artifact_relationships`, and `get_repository_ontology` gives a repository overview. `query_artifact_metadata` runs read-only metadata analytics and `get_data_sources` lists indexed repositories. The slow `chat` synthesizes an answer over the codebase and is called on explicit request. The connection is remote: `https://mcp.codealive.ai/api` with an `Authorization: Bearer` header and a key from the web app.

Who it is for. Developers of large monorepos where the built-in agent search falls short and reading files one by one is slow and expensive in tokens.

Good fit when

  • The agent must find where user service authentication is handled without knowing file names
  • An exact match of a JWT regex is needed across dozens of packages
  • Before a refactor it must be shown who calls a function and which classes inherit it
  • Indexed repository data and metadata are needed for analytics across several projects

Not a fit when

  • The repository is small: the built-in agent search is enough without an extra service
  • Code must not go to an external service: repositories are indexed in CodeAlive
  • Only text search over local files is needed: `grep` is simpler

Example request

Find in the indexed repository where the authorization token is validated and show which functions call that code

MCP

Transport
http
Authentication
API key

Environment variables

CODEALIVE_API_KEY required, secret
API key from the MCP & API section at app.codealive.ai, needed for a local run.

Limitations

A CodeAlive account and an API key are required: it is created in the MCP & API section at app.codealive.ai and shown once. Repositories must be indexed in the service, otherwise there is nothing to search. The main mode is remote and local run is for development, with `CODEALIVE_API_KEY` set in the process environment. The heavy `chat` works longer than plain search, and agents are advised to start with `semantic_search` and `grep_search`. For the agent to choose CodeAlive over built-in search, rules are added to `AGENTS.md` or `CLAUDE.md`.

How to disable. Remove the CodeAlive server from the client MCP config and revoke the API key in the MCP & API section at app.codealive.ai.

FAQ

How do I connect CodeAlive MCP to Claude Code?

Create a key at app.codealive.ai in the MCP & API section and add the remote server `https://mcp.codealive.ai/api` following the Claude Code guide on docs.codealive.ai.

What authentication does the server use?

An `Authorization: Bearer` header with an API key over Streamable HTTP.

Is there an agent skill?

Yes, `npx skills add CodeAlive-AI/codealive-skills@codealive-context-engine`, and Claude Code has a plugin in the `CodeAlive-AI/codealive-skills` marketplace.

Can I self-host it?

Yes, the server runs in HTTP mode with `python src/codealive_mcp_server.py --transport http`, details are in the Self-Hosting docs section.

What license applies?

MIT.

Discussion

No questions or reviews yet.

An MCP server that turns a large codebase into a semantic feature graph for precise navigation and meaning-based search

MCP serverExtended access2KRepository stars

An MCP server that builds a code graph of functions, classes, calls and routes and answers the agent's structural queries locally

MCP serverExtended access48.5KRepository stars

An MCP server that indexes code with tree-sitter and returns individual functions and classes to the agent instead of whole files

MCP serverExtended access2.8KRepository stars

A local codebase index for agents: hybrid search, dependency graphs and change impact analysis

MCP serverExtended access3.4KRepository stars
Official

A semantic code search MCP server: indexes the project into a vector database and gives the agent only relevant snippets

MCP serverExtended access12.7KRepository stars
Official

A local code graph with version history: an MCP server that tells the agent who calls a function and what an edit breaks

MCP serverExtended access516Repository stars

README in short

A description of CodeAlive and a list of 11 tools with examples, an agent skill, a quick start with the remote server and a key, client guide links, local development and HTTP deployment.

README badge

Report a problem