# ChromaDB Remote MCP Server

A remote MCP server connects AI clients to ChromaDB over Streamable HTTP for semantic search and vector database operations.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 12
- Author: meloncafe
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/meloncafe/chromadb-remote-mcp
- Updated: 2026-10-24
- HTML: https://affpapa.org/en/nevibecode/mcp/chromadb-remote-mcp-server

## What it does

ChromaDB Remote MCP Server exposes a self-hosted ChromaDB instance to Claude Code and other MCP-compatible clients. It receives MCP requests, maps them to ChromaDB operations, and keeps vector embeddings available for semantic search. The same database can serve clients on different devices.

The server combines the MCP protocol with an authentication gateway and a REST API proxy. Python and JavaScript clients can reach ChromaDB through `/api/v2/*`; `/docs` and `/openapi.json` expose the API documentation. Clients connect through the `/mcp` endpoint, while Claude Code uses the `claude mcp add` command.

A manual Docker setup requires an `.env` file with `MCP_AUTH_TOKEN`, `PORT`, and `CHROMA_DATA_PATH`, followed by Docker Compose. Local development uses `yarn install`, `yarn build`, and `yarn start`. The server uses Streamable HTTP; its README marks SSE transport as deprecated.

## Requirements
- Needs: Docker
- Remote server: `https://meloncafe.github.io/chromadb-remote-mcp/mcp`
- Environment variables: `CHROMADB_MCP_URL`, `MCP_AUTH_TOKEN` — keys and tokens: ask the human for the values, never commit them or put them in shared files.

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http chromadb-remote-mcp https://meloncafe.github.io/chromadb-remote-mcp/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add chromadb-remote-mcp-server --url https://meloncafe.github.io/chromadb-remote-mcp/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "chromadb-remote-mcp-server": {
            "url": "https://meloncafe.github.io/chromadb-remote-mcp/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "chromadb-remote-mcp-server": {
            "type": "http",
            "url": "https://meloncafe.github.io/chromadb-remote-mcp/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http chromadb-remote-mcp-server https://meloncafe.github.io/chromadb-remote-mcp/mcp
```

## When to use it
- connect Claude Code to a remote ChromaDB instance
- search stored content with vector semantics
- share one database across AI clients
- expose ChromaDB to Python and JavaScript through REST

## Example prompts
- Connect this MCP server and search ChromaDB for information related to the current task.
- Store the project context in ChromaDB so we can retrieve it in a later request.
- Check the remote ChromaDB connection and run a semantic search over the stored data.

## Similar
- [cli-anything-chromadb](https://affpapa.org/en/nevibecode/skills/cli-anything-chromadb.md) — skill · 505 installs
- [rag-engineer](https://affpapa.org/en/nevibecode/skills/sickn33-rag-engineer.md) — skill · 1000 installs
- [longhand](https://affpapa.org/en/nevibecode/plugins/longhand.md) — plugin · ★ 14
- [rag-implementation](https://affpapa.org/en/nevibecode/skills/rag-implementation.md) — skill · 13906 installs
- [similarity-search-patterns](https://affpapa.org/en/nevibecode/skills/similarity-search-patterns.md) — skill · 10700 installs
- [upstash-vector-js](https://affpapa.org/en/nevibecode/skills/upstash-vector-js.md) — skill · 1300 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md