Skip to content
Back to search

codealive-mcp

Context engine for large codebases, exposed through MCP. Gives AI coding agents precise repository context; benchmarked at frontier-agent quality with ~25x lower model cost and 45% fewer tokens with semantic search.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/CodeAlive-AI/codealive-mcp with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/CodeAlive-AI/codealive-mcp'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/codealive-ai--codealive-mcp/badge.svg)](/listings/codealive-ai--codealive-mcp)

HTML

<a href="/listings/codealive-ai--codealive-mcp"><img src="/api/listings/codealive-ai--codealive-mcp/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

**The fastest way to get started** - no installation required! Our remote MCP server at `https://mcp.codealive.ai/api` provides instant access to CodeAlive's capabilities.

Step 1: Get Your API Key

1. Sign up at [https://app.codealive.ai/](https://app.codealive.ai/) 2. Navigate to **MCP & API** 3. Click **"+ Create API Key"** 4. Copy your API key immediately - you won't see it again!

Step 2: Open Your Client Guide

Choose your client in the [MCP integration guides](https://docs.codealive.ai/integrations/mcp) and follow the current setup instructions there.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 413 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # CodeAlive MCP: Deepest Context Engine for your projects (especially for large codebases)
  2. <!-- MCP Server Name: io.github.codealive-ai.codealive-mcp -->
  3. [![CodeAlive Logo](https://app.codealive.ai/images/logos/dark-logo.svg)](https://www.codealive.ai/)
  4. **Connect your AI assistant to CodeAlive's powerful code understanding platform in seconds!**
  5. This MCP (Model Context Protocol) server enables AI clients like Claude Code, Cursor, Claude Desktop, Continue, VS Code (GitHub Copilot), Cline, Codex, OpenCode, SourceCraft Code Assistant, SourceCraft CLI, Zed, KodaCode, GigaCode, Qwen Code, Gemini CLI, Roo Code, Goose, Kilo Code, Windsurf, Kiro, Qoder, n8n, and Amazon Q Developer to access CodeAlive's advanced semantic code search and codebase interaction features.
  6. ## What is CodeAlive?
  7. CodeAlive is a Context Engine for large codebases, powered by graph-based retrieval and exposed through MCP. It gives AI agents like Cursor, Claude Code, Codex, and other MCP-compatible tools precise repository context instead of forcing them to read files blindly. In our RepoQA benchmark, CodeAlive + Qwen3.6 deep reached frontier-agent quality at ~25x lower model cost, and semantic search reduced captured tokens by 45%.
  8. It's like Context7, but for your (large) codebases.
  9. It allows AI-Coding Agents to:
  10. * **Find relevant code faster** with semantic search
  11. * **Understand the bigger picture** beyond isolated files
  12. * **Provide better answers** with full project context
  13. * **Reduce costs and time** by removing guesswork
  14. ## 🛠 Available Tools
  15. Once connected, you'll have access to these powerful tools:
  16. 1. **`get_data_sources`** - List your indexed repositories and workspaces
  17. 2. **`semantic_search`** - Canonical semantic search across indexed artifacts
  18. 3. **`grep_search`** - Exact literal or regex text search inside file content, plus literal file-name/path matching (returns files like `Form.xml` even when their content never mentions the name), with line-level previews for content matches
  19. 4. **`get_repository_ontology`** - Get repository-level orientation for one selected repository
  20. 5. **`get_file_tree`** - Inspect a bounded file tree for one repository
  21. 6. **`read_file`** - Read a repository-relative file path, optionally with a line range
  22. 7. **`fetch_artifacts`** - Load the full source for relevant search hits (missing or inaccessible identifiers are reported back, not silently dropped)
  23. 8. **`get_artifact_relationships`** - Expand call graph, inheritance, and reference relationships for one artifact
  24. 9. **`get_artifact_query_schema`** - Inspect supported ArtifactQuery entities, fields, and examples
  25. 10. **`query_artifact_metadata`** - Run read-only metadata analytics across selected repositories
  26. 11. **`chat`** - Stateless, slower synthesized codebase Q&A; call only when explicitly requested
  27. ## 🎯 Usage Examples
  28. After setup, try these commands with your AI assistant:
  29. - *"Show me all available repositories"* → Uses `get_data_sources`
  30. - *"Find authentication code in the user service"* → Uses `semantic_search`
  31. - *"Find the exact regex that matches JWT tokens"* → Uses `grep_search`
  32. - *"Explain how the payment flow works in this codebase"* → Usually starts with `semantic_search`/`grep_search`, then optionally uses `chat`
  33. `semantic_search` and `grep_search` should be the default tools for most agents. `chat` is a slower stateless synthesis fallback that can take substantially longer than retrieval, and is usually unnecessary when an agent can run a multi-step workflow with ontology, search, fetch/read, relationships, ArtifactQuery, and local file reads. If your agent supports subagents, the highest-confidence path is to delegate a focused subagent that orchestrates `semantic_search` and `grep_search` first.
  34. ## 📚 Agent Skill
  35. For an even better experience, install the [CodeAlive Agent Skill](https://github.com/CodeAlive-AI/codealive-skills) alongside the MCP server. The MCP server gives your agent access to CodeAlive's tools; the skill teaches it the best workflows and query patterns to use them effectively.
  36. **For most agents** (Cursor, Copilot, Gemini CLI, Codex, and [30+ others](https://agentskills.io/)) — install the skill:
  37. ```bash
  38. npx skills add CodeAlive-AI/codealive-skills@codealive-context-engine
  39. ```
  40. **For Claude Code** — install the plugin (recommended), which includes the skill plus Claude-specific enhancements:
  41. ```
  42. /plugin marketplace add CodeAlive-AI/codealive-skills
  43. /plugin install codealive@codealive-marketplace
  44. ```
  45. ## Table of Contents
  46. * [Agent Skill](#-agent-skill)
  47. * [Quick Start (Remote)](#-quick-start-remote)
  48. * [AI Client Integrations](#-ai-client-integrations)
  49. * [Advanced: Local Development](#-advanced-local-development)
  50. * [Community Plugins](#-community-plugins)
  51. * [HTTP Deployment (Self-Hosted & Cloud)](#-http-deployment-self-hosted--cloud)
  52. * [Windows & WSL](#-windows--wsl)
  53. * [Available Tools](#-available-tools)
  54. * [Usage Examples](#-usage-examples)
  55. * [Troubleshooting](#-troubleshooting)
  56. * [Publishing to MCP Registry](#-publishing-to-mcp-registry)
  57. * [License](#-license)
  58. ## 🚀 Quick Start (Remote)
  59. **The fastest way to get started** - no installation required! Our remote MCP server at `https://mcp.codealive.ai/api` provides instant access to CodeAlive's capabilities.
  60. ### Step 1: Get Your API Key
  61. 1. Sign up at [https://app.codealive.ai/](https://app.codealive.ai/)
  62. 2. Navigate to **MCP & API**
  63. 3. Click **"+ Create API Key"**
  64. 4. Copy your API key immediately - you won't see it again!
  65. ### Step 2: Open Your Client Guide
  66. Choose your client in the [MCP integration guides](https://docs.codealive.ai/integrations/mcp) and follow the current setup instructions there.
  67. ## 🚀 Quick Start (Agentic Installation)
  68. You may ask your AI agent to install the CodeAlive MCP server for you.
  69. 1. Copy-paste the following prompt into your AI agent. Do not include your API key in the prompt:
  70. ```
  71. Add the CodeAlive MCP server by following the guide for my client at https://docs.codealive.ai/integrations/mcp
  72. Prefer the Remote HTTP option when available. Do not ask me to paste an API key into chat. When the key is needed, ask me to create a CodeAlive API key and copy it to my clipboard. After I confirm, insert it directly from the clipboard into the required secure configuration without displaying, echoing, logging, or exposing it in command arguments, command output, or model context. If you cannot safely use the clipboard without exposing the value, tell me exactly where to paste it myself.
  73. ```
  74. Then allow execution.
  75. 2. Restart your AI agent.
  76. ## 🤖 AI Client Integrations
  77. Client-specific configuration is maintained in the CodeAlive documentation so file paths, transports, and authentication guidance stay current.
  78. **Start here:** [MCP integration guides](https://docs.codealive.ai/integrations/mcp)
  79. | Client | Setup guide |
  80. |---|---|
  81. | Claude Code | [Claude Code](https://docs.codealive.ai/integrations/mcp/claude-code) |
  82. | Claude Desktop | [Claude Desktop](https://docs.codealive.ai/integrations/mcp/claude-desktop) |
  83. | Cursor | [Cursor](https://docs.codealive.ai/integrations/mcp/cursor) |
  84. | Visual Studio Code | [VS Code](https://docs.codealive.ai/integrations/mcp/vscode) |
  85. | Windsurf | [Windsurf](https://docs.codealive.ai/integrations/mcp/windsurf) |
  86. | Cline | [Cline](https://docs.codealive.ai/integrations/mcp/cline) |
  87. | Continue | [Continue](https://docs.codealive.ai/integrations/mcp/continue) |
  88. | Codex | [Codex](https://docs.codealive.ai/integrations/mcp/codex) |
  89. | Gemini CLI | [Gemini CLI](https://docs.codealive.ai/integrations/mcp/gemini-cli) |
  90. | Amazon Q Developer | [Amazon Q](https://docs.codealive.ai/integrations/mcp/amazon-q) |
  91. | OpenCode | [OpenCode](https://docs.codealive.ai/integrations/mcp/opencode) |
  92. | SourceCraft Code Assistant and SourceCraft CLI | [SourceCraft](https://docs.codealive.ai/integrations/mcp/sourcecraft) |
  93. | Zed | [Zed](https://docs.codealive.ai/integrations/mcp/zed) |
  94. | ChatGPT | [ChatGPT](https://docs.codealive.ai/integrations/mcp/chatgpt) |
  95. | OpenClaw | [OpenClaw](https://docs.codealive.ai/integrations/mcp/openclaw) |
  96. | KodaCode, GigaCode, Roo Code, Goose, Kilo Code, Qwen Code, Kiro, Qoder, JetBrains AI Assistant, n8n, and more | [Other agents](https://docs.codealive.ai/integrations/mcp/other-agents) |
  97. For an unlisted client, use these generic connection details and adapt them to the client's MCP configuration format:
  98. - **Endpoint:** `https://mcp.codealive.ai/api`
  99. - **Transport:** Streamable HTTP
  100. - **Authentication header:** `Authorization: Bearer YOUR_API_KEY_HERE`
  101. For a private deployment, replace the endpoint with your server's `/api` URL. See [Self-Hosting](https://docs.codealive.ai/integrations/mcp/self-hosting) for deployment guidance.
  102. > **Connecting the server is half the setup.** Coding agents may continue using their built-in search unless project instructions tell them to prefer CodeAlive. Ready-made rules for `AGENTS.md`, `CLAUDE.md`, and client-specific instruction files are in [Instructing Coding Agents](https://docs.codealive.ai/guides/instructing-agents).
  103. ---
  104. ## 🔧 Advanced: Local Development
  105. **For developers who want to customize or contribute to the MCP server.**
  106. ### Prerequisites
  107. * Python 3.11+
  108. * [uv](https://github.com/astral-sh/uv) (recommended) or pip
  109. ### Installation
  110. ```bash
  111. # Clone the repository
  112. git clone https://github.com/CodeAlive-AI/codealive-mcp.git
  113. cd codealive-mcp
  114. # Setup with uv (recommended)
  115. uv venv
  116. source .venv/bin/activate # Windows: .venv\Scripts\activate
  117. uv pip install -e .
  118. # Or setup with pip
  119. python -m venv .venv
  120. source .venv/bin/activate # Windows: .venv\Scripts\activate
  121. pip install -e .
  122. ```
  123. ### Local Server Configuration
  124. After installing the server locally, point your MCP client at `.venv/bin/python` with `src/codealive_mcp_server.py` as the first argument and provide `CODEALIVE_API_KEY` in the process environment. Client-specific configuration belongs in the [MCP integration guides](https://docs.codealive.ai/integrations/mcp).
  125. ### Running HTTP Server Locally
  126. ```bash
  127. # Start local HTTP server
  128. export CODEALIVE_API_KEY="your_api_key_here"
  129. python src/codealive_mcp_server.py --transport http --host localhost --port 8000
  130. # Test health endpoint
  131. curl http://localhost:8000/health
  132. ```
  133. HTTP transport validates `Host` and browser `Origin` headers. Loopback hosts
  134. (`localhost`, `127.0.0.1`, `::1`) work without extra configuration. For a
  135. shared hostname, configure an exact allowlist:
  136. ```bash
  137. export CODEALIVE_MCP_ALLOWED_HOSTS="mcp.codealive.yourcompany.com"
  138. # Only for browser callers; ordinary MCP clients do not send Origin.
  139. export CODEALIVE_MCP_ALLOWED_ORIGINS="https://mcp.codealive.yourcompany.com"
  140. python src/codealive_mcp_server.py --transport http --host 0.0.0.0 --port 8000
  141. ```
  142. The equivalent repeatable CLI options are `--allowed-host` and
  143. `--allowed-origin`. Do not use `*` for an Internet-facing server.
  144. ### Testing Your Local Installation
  145. After making changes, quickly verify everything works:
  146. ```bash
  147. # Match pyproject.toml exactly; older uv versions reject the locked setup.
  148. uv --version # expected: uv 0.11.28
  149. uv sync --locked --extra test
  150. # Install the repository pre-push dependency audit once per clone
  151. ./scripts/setup-hooks.sh
  152. # Quick smoke test (recommended)
  153. make smoke-test
  154. # Or run directly
  155. python smoke_test.py
  156. # With your API key for full testing
  157. CODEALIVE_API_KEY=your_key python smoke_test.py
  158. # Run unit tests
  159. make unit-test
  160. # Run all tests
  161. make test
  162. # Equivalent direct locked test run
  163. uv run pytest src/tests/ -q
  164. ```
  165. The smoke test verifies:
  166. - Server starts and connects correctly
  167. - All tools are registered
  168. - Each tool responds appropriately
  169. - Parameter validation works
  170. - Runs in ~5 seconds
  171. ---
  172. ## 🌐 Community Plugins
  173. - [Gemini CLI — CodeAlive Extension](https://github.com/akolotov/gemini-cli-codealive-extension)
  174. - [Gemini CLI setup guide](https://docs.codealive.ai/integrations/mcp/gemini-cli)
  175. ---
  176. ## 🚢 HTTP Deployment (Self-Hosted & Cloud)
  177. **Deploy the MCP server as an HTTP service for team-wide access or integration with self-hosted CodeAlive instances.**
  178. ### Deployment Options
  179. The CodeAlive MCP server can be deployed as an HTTP service using Docker. This allows multiple AI clients to connect to a single shared instance, and enables integration with self-hosted CodeAlive deployments.
  180. ### Docker Compose (Recommended)
  181. Create a `docker-compose.yml` file based on our example:
  182. ```bash
  183. # Download the example
  184. curl -O https://raw.githubusercontent.com/CodeAlive-AI/codealive-mcp/main/docker-compose.example.yml
  185. mv docker-compose.example.yml docker-compose.yml
  186. # Edit configuration (see below)
  187. nano docker-compose.yml
  188. # Start the service
  189. docker compose up -d
  190. # Check health
  191. curl http://localhost:8000/health
  192. ```
  193. **Configuration Options:**
  194. 1. **For CodeAlive Cloud (default):**
  195. - Remove `CODEALIVE_BASE_URL` environment variable (uses default `https://app.codealive.ai`)
  196. - For remote clients with OAuth support, configure only `https://mcp.codealive.ai/api` and
  197. complete the browser sign-in when prompted
  198. - Existing API-key clients remain supported via `Authorization: Bearer YOUR_KEY`
  199. 2. **For Self-Hosted CodeAlive:**
  200. - Set `CODEALIVE_BASE_URL` to your CodeAlive instance URL (e.g., `https://codealive.yourcompany.com`)
  201. - Set `CODEALIVE_MCP_ALLOWED_HOSTS` to the exact hostname clients use for this MCP server
  202. - Clients must provide their API key via `Authorization: Bearer YOUR_KEY` header
  203. See `docker-compose.example.yml` for the complete configuration template.
  204. For example, current Codex and Claude Code clients can use browser OAuth without storing a
  205. CodeAlive API key:
  206. ```bash
  207. codex mcp add codealive --url https://mcp.codealive.ai/api
  208. codex mcp login codealive
  209. claude mcp add --transport http codealive https://mcp.codealive.ai/api
  210. # Start Claude Code and run /mcp to authenticate.
  211. ```
  212. Cursor and OpenCode also discover OAuth automatically from the same URL. Use
  213. `cursor-agent mcp login codealive` or `opencode mcp auth codealive` when their UI does not prompt
  214. automatically. API-key configuration remains available as a compatibility option.
  215. ### OAuth 2.1 deployment profile
  216. Remote HTTP deployments can enable browser authorization while keeping legacy API-key clients
  217. working during rollout. OAuth mode publishes MCP Protected Resource Metadata, validates exact
  218. issuer/resource-bound JWTs, and exchanges them for a separate short-lived Tool API token. The
  219. incoming MCP bearer token is never forwarded downstream.
  220. | Environment variable | Purpose |
  221. |---|---|
  222. | `CODEALIVE_MCP_OAUTH_ENABLED=true` | Enables OAuth validation and MCP authorization discovery for HTTP transport |
  223. | `CODEALIVE_OAUTH_ISSUER` | Exact OpenIddict issuer, with a trailing slash |
  224. | `CODEALIVE_MCP_RESOURCE` | Exact public MCP resource URL; its path is also the HTTP MCP path |
  225. | `CODEALIVE_TOOL_API_RESOURCE` | Downstream audience; defaults to `urn:codealive:tool-api` |
  226. | `CODEALIVE_OAUTH_INTERNAL_CLIENT_ID` | Confidential resource-server client used only for token exchange |
  227. | `CODEALIVE_OAUTH_INTERNAL_CLIENT_SECRET` | Required secret for that internal client; startup fails closed when it is missing |
  228. The authorization server and MCP service values must match exactly. In CodeAlive Web.Server the
  229. corresponding settings live under `McpOAuth` (`Enabled`, `Issuer`, `Resource`,
  230. `ToolApiResource`, `InternalClientId`, and `InternalClientSecret`). Persist the Web.Server Data
  231. Protection key ring and OpenIddict signing/encryption certificates across replicas and restarts.
  232. For a zero-downtime internal credential rotation, give the new credential a new client ID, deploy
  233. Web.Server with both current and `PreviousInternalClientId`/`PreviousInternalClientSecret`, roll
  234. MCP replicas to the new current pair, then remove the previous pair. Web.Server deliberately fails
  235. startup instead of changing a secret in place under an existing client ID.
  236. Enable the Web.Server and MCP flags in the same rollout; a half-enabled deployment is not a valid
  237. steady state. API-key credentials retain their explicit legacy grammar and are never used as a
  238. fallback after OAuth validation fails.
  239. ### Connecting MCP Clients to Your Deployed Instance
  240. Use the same generic connection details as CodeAlive Cloud, replacing the endpoint with your deployment's `/api` URL:
  241. - **Endpoint:** `https://your-server.example.com/api`
  242. - **Transport:** Streamable HTTP
  243. - **Authentication header:** `Authorization: Bearer YOUR_API_KEY_HERE`
  244. For the exact configuration format, open the relevant [client integration guide](https://docs.codealive.ai/integrations/mcp).
  245. ## 🪟 Windows & WSL
  246. Use the client-specific documentation for Windows and WSL setup:
  247. - [Claude Code](https://docs.codealive.ai/integrations/mcp/claude-code)
  248. - [Claude Desktop](https://docs.codealive.ai/integrations/mcp/claude-desktop)
  249. - [All MCP integration guides](https://docs.codealive.ai/integrations/mcp)
  250. - [Troubleshooting](https://docs.codealive.ai/troubleshooting)
  251. For self-hosted servers running in WSL2, Windows clients must be able to reach the server's `/api` endpoint. Use mirrored networking on supported Windows 11 versions or connect through the WSL2 VM address.
  252. ## 🐞 Troubleshooting
  253. ### Quick Diagnostics
  254. 1. **Test the hosted service:**
  255. ```bash
  256. curl https://mcp.codealive.ai/health
  257. ```
  258. 2. **Check your API key:**
  259. ```bash
  260. curl -H "Authorization: Bearer YOUR_API_KEY" https://app.codealive.ai/api/v1/data_sources
  261. ```
  262. 3. **Enable debug logging:** Add `--debug` to local server args
  263. ### Common Issues
  264. - **"Connection refused"** → Check internet connection
  265. - **"401 Unauthorized"** → Verify your API key
  266. - **"No repositories found"** → Check API key permissions in CodeAlive dashboard
  267. - **Client-specific logs** → See your AI client's documentation for MCP logs
  268. ### Windows / WSL Issues
  269. - **`docker: command not found` in WSL** → Enable Docker Desktop WSL integration for your distro (Settings → Resources → WSL integration), or use the full path `/usr/bin/docker`
  270. - **`ENOENT` or `spawn error` for `npx`/`python`** → Non-interactive WSL shells don't inherit `nvm`/`pyenv` paths. Use absolute paths in MCP configs
  271. - **`Connection refused` to self-hosted server in WSL2** → WSL2 uses NAT networking; `localhost` differs between Windows and WSL2. Enable mirrored networking in `.wslconfig` or use the WSL2 VM IP (`hostname -I`)
  272. - **Claude Desktop can't connect to WSL MCP server** → Claude Desktop doesn't support WSL subprocess spawning. Use Remote HTTP (`https://mcp.codealive.ai/api`), Docker Desktop, or the `wsl.exe` proxy pattern (see [Windows & WSL section](#-windows--wsl))
  273. ### Getting Help
  274. - 📧 Email: support@codealive.ai
  275. - 🐛 Issues: [GitHub Issues](https://github.com/CodeAlive-AI/codealive-mcp/issues)
  276. ---
  277. ## 📦 Publishing to MCP Registry
  278. For maintainers: see [DEPLOYMENT.md](DEPLOYMENT.md) for instructions on publishing new versions to the MCP Registry.
  279. ---
  280. ## Privacy Policy
  281. CodeAlive processes the repositories and queries you send through this extension in order to provide semantic search and codebase analysis. For complete privacy details, see [CodeAlive Privacy Policy](https://www.codealive.ai/privacy/).
  282. ---
  283. ## 📄 License
  284. MIT License - see [LICENSE](LICENSE) file for details.
  285. ---
  286. **Ready to supercharge your AI assistant with deep code understanding?**
  287. [Get started now →](https://app.codealive.ai/)

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.