Supavec

Supavec is an MCP server that enables AI assistants to search and retrieve embeddings and indexed file content from the Supavec vector platform. Built for software engineers and knowledge workers using MCP-compatible environments like Claude Desktop, Cursor, and VS Code Copilot, the server provides direct access to remote document repositories. Instead of manually copying context or configuring custom retrieval pipelines, users can invoke Supavec tools directly from their chat interface. The server connects to the Supavec API using an authentication key provided either via command-line arguments or environment variables. Once connected, clients can query specific documents to extract semantically relevant text passages or list existing files stored in their Supavec account. This retrieval architecture supports document-grounded question answering, contextual code navigation, and technical documentation search without requiring local embedding generation or manual data preprocessing.

Category: AI Memory & Context

Tags: embeddings, rag, supavec, vector-search

Visit Supavec

How to install and configure Supavec

1. Obtain an API key from your Supavec account settings. 2. Add the server configuration to your MCP client config file (such as .cursor/mcp.json, .vscode/mcp.json, or Claude Desktop configuration): json { "mcpServers": { "supavec": { "command": "npx", "args": [ "-y", "@supavec/mcp-server@latest", "--api-key", "your_supavec_api_key" ] } } } 3. Alternatively, export the environment variable export SUPAVEC_API_KEY=your_supavec_api_key and run without the --api-key flag. 4. Restart your MCP client to load the tools.

What you can do with Supavec

• Search indexed documentation files for setup guides and API endpoints directly inside Claude or Cursor. - Retrieve semantically relevant passages from user-uploaded specifications to answer architectural and implementation questions accurately. - Query existing enterprise knowledge files in Supavec without requiring local vector database instances. - List uploaded team documents and files to inspect available resources, file IDs, and upload timestamps. - Ground development answers in VS Code Copilot using embeddings fetched dynamically from hosted Supavec files.

Key facts

• https://github.com/supavec/mcp-server
• AI Memory & Context, Databases & Data Stores
• embeddings, rag, supavec, vector-search

Part of MCP Servers

How do I install Supavec?

You can run it without manual installation using npx by executing npx @supavec/mcp-server --api-key your_api_key. Alternatively, install it globally using npm install -g @supavec/mcp-server and run supavec-mcp with your key.

What can Supavec do?

The server exposes tools to retrieve context from Supavec. Specifically, fetch-embeddings searches a specified file by ID and query to return relevant content, while list-user-files retrieves user files with pagination and sorting controls.

Which MCP clients work with Supavec?

Supavec works with any Model Context Protocol compliant client. The documentation provides explicit setup configurations for Cursor, VS Code with Copilot, and Claude Desktop.

How do I authenticate the Supavec server?

You can authenticate by passing your API key via the --api-key command-line argument or by setting the SUPAVEC_API_KEY environment variable. The command-line argument takes precedence if both are set.

Is Supavec open source?

Yes, the Supavec MCP server is open source and licensed under the MIT License on GitHub.