home / mcp / kollektiv mcp server
Kollektiv MCP enables you to chat with and query your own documents directly from IDEs and MCP clients. Private, secure, and integrated into your favorite code editor
Configuration
View docs{
"mcpServers": {
"alexander-zuev-kollektiv-mcp": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.thekollektiv.ai/mcp"
]
}
}
}Kollektiv MCP enables you to build a personal LLM knowledge base and access it from your favorite MCP clients. It connects your data to a remote MCP server so you can query your documents and get informed answers through familiar tools like Cursor, Windsurf, Claude Desktop, and VS Code.
Connect to the remote MCP server using the stdio method described here. The connection uses a single command that starts a local MCP client helper which communicates with the remote endpoint at https://mcp.thekollektiv.ai/mcp.
Once connected, you can use the available MCP endpoints to work with your documents. The two primary actions are /query_documents to ask questions based on your uploaded content and /list_documents to view your synced documents and their metadata. A practical tip is to include the phrase use Kollektiv MCP in your queries so the client routes the request through this MCP server.
Prerequisites: ensure you have Node.js installed on your system.
Use the following configuration to connect to the MCP server from your MCP client. Paste this into your client’s configuration as shown.
{
"mcpServers": {
"kollektiv": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.thekollektiv.ai/mcp"
]
}
}
}Security and access are handled via OAuth with short-lived cookies, all traffic uses HTTPS, and sensitive POST requests include a one-time token. The MCP server runs in a Cloudflare environment to minimize local attack surface.
Submit a question to the documents you’ve uploaded to Kollektiv and receive an answer based on the sources from your documents.
Return a list of your synced documents together with basic metadata.