Trieve - MCP Server

trieve-mcp-server

A Model Context Protocol (MCP) server that provides agentic tools for interacting with the Trieve API. This server enables AI agents to search and interact with Trieve datasets through a standardized interface.

Features

Search across Trieve datasets using semantic search
List and access dataset information
Support for both environment variables and command-line arguments
Built with TypeScript for type safety and better developer experience

Installation

Installing via Smithery

To install Trieve for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install trieve-mcp-server --client claude

Manual Installation

npm install trieve-mcp-server

Configuration

Copy the .env.dist file to .env and fill in your Trieve credentials:

cp .env.dist .env

Required environment variables:

TRIEVE_API_KEY: Your Trieve API key from dashboard.trieve.ai
TRIEVE_ORGANIZATION_ID: Your Trieve organization ID from dashboard.trieve.ai

Optional environment variables:

TRIEVE_DATASET_ID: Specific dataset ID to use (if not provided via CLI)

Command-line arguments (override environment variables):

trieve-mcp-server --api-key <your-api-key> --org-id <your-org-id> [--dataset-id <dataset-id>]

Usage

Starting the Server

trieve-mcp-server

Available Tools

search

Search through a specified Trieve dataset.

Parameters:

query (string): The search query
datasetId (string): ID of the dataset to search in
searchType (string, optional): "semantic" (default), "fulltext", "hybrid", or "bm25"
filters (object, optional): Advanced filtering options
highlightOptions (object, optional): Customize result highlighting
page (number, optional): Page number, default 1
pageSize (number, optional): Results per page, default 10

Example:

{
  "query": "example search query",
  "datasetId": "your-dataset-id",
  "searchType": "semantic",
  "page": 1,
  "pageSize": 10
}

Available Resources

The server exposes Trieve datasets as resources with the following URI format:

trieve://datasets/{dataset-id}

Usage with Claude Desktop

The Trieve MCP Server supports MCP integration with Claude Desktop. Place the following in your Claude Desktop's claude_desktop_config.json.

{
  "mcpServers": {
    "trieve-mcp-server": {
      "command": "npx",
      "args": ["trieve-mcp-server@latest"],
      "env": {
        "TRIEVE_API_KEY": "$TRIEVE_API_KEY",
        "TRIEVE_ORGANIZATION_ID": "$TRIEVE_ORGANIZATION_ID",
        "TRIEVE_DATASET_ID": "$TRIEVE_DATASET_ID"
      }
    }
  }
}

Note: Instead of environment variables, --api-key, --org-id, and --dataset-id can be used as command-line arguments.

Once Claude Desktop starts, attachments will be available that correspond to the datasets available to the Trieve organization. These can be used to select a dataset. After that, begin chatting with Claude and ask for information about the dataset. Claude will use search as needed in order to filter and break down queries, and may make multiple queries depending on your task.

Development

Setup

Clone the repository
Install dependencies:

npm install

Copy .env.dist to .env and configure your credentials
Build the project:

npm run build

Scripts

npm run build: Build the TypeScript project
npm run watch: Watch for changes and rebuild
npm run test: Run tests
npm run inspector: Run the MCP inspector for debugging

License

MIT