RequestHuntRequestHunt

MCP Integration

Use RequestHunt with Claude Desktop, Cursor, or any MCP-compatible AI client. The Model Context Protocol allows AI assistants to access external tools and data sources.

Quick Start

  1. Get an API key from your dashboard
  2. Add the configuration below to your MCP client
  3. Restart your client and start asking about feature requests

Installation

The MCP server is available as an npm package. No manual installation required - your MCP client will automatically download it via npx.

Optional: Global Installation
npm install -g @resciencelab/requesthunt-mcp-server

Claude Desktop

Add to your Claude Desktop configuration file:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "requesthunt": {
      "command": "npx",
      "args": ["-y", "@resciencelab/requesthunt-mcp-server"],
      "env": {
        "REQUESTHUNT_API_KEY": "rh_live_your_api_key_here"
      }
    }
  }
}

Cursor

Add to your Cursor MCP settings (Settings → MCP → Add Server):

Cursor MCP Configuration
{
  "mcpServers": {
    "requesthunt": {
      "command": "npx",
      "args": ["-y", "@resciencelab/requesthunt-mcp-server"],
      "env": {
        "REQUESTHUNT_API_KEY": "rh_live_your_api_key_here"
      }
    }
  }
}

Available Tools

search_requests

Full-text search across all feature requests with optional AI topic expansion.

ParameterTypeDescription
querystringSearch query (required)
limitnumberMax results (1-100, default: 20)
expandbooleanEnable AI expansion + realtime scraping
platformsarrayFilter: reddit, x, github

list_requests

List feature requests with filters and pagination.

ParameterTypeDescription
topicstringFilter by topic (e.g., "Notion")
categorystringFilter by category slug
platformstringFilter: x, reddit, github
sourcestringcached or realtime
sortBystringnew or top

get_topics

List available topics organized by category. Use these to filter feature requests. No parameters required.

Example Prompts

Search for feature requests:

"Search for feature requests about dark mode"

Find requests for a specific product:

"Find what users are requesting for Notion"

With AI expansion for more results:

"Search for 'offline mode' feature requests with expand enabled"

Resources