Model Context Protocol

Connect PainSpotter to Claude, Cursor & any LLM

Query AI-scored business opportunities right inside your IDE via the Model Context Protocol (MCP) — or pull them through our REST API.

Get your API key →REST API Docs →View on Glama ↗View on Smithery ↗

What is the PainSpotter MCP Server

PainSpotter mines Reddit, Hacker News and Product Hunt discussions and uses AI to surface validated business opportunities with commercial scores. The MCP server exposes that data to any MCP-aware client — Claude Desktop, Cursor, Claude Code, VS Code — so an LLM can search opportunities, inspect commercial scores and pull trending pain points without leaving your workflow. Data is layered: Category (domain) → Theme (a clustered pain point with trend signal) → Opportunity (a concrete product idea).

Tools

Seven tools on the hosted endpoint: four free (including blog, zero quota cost) and three Pro-only search/theme tools.

Tool	What it does	Tier
`get_overview`	Global landscape: categories with theme/opportunity counts + a trending snapshot.	Free
`get_opportunity`	Full detail of one opportunity: score breakdown, MVP features, competitors, risks (+ related blog URL).	Free
`list_blog_posts`	Recent published weekly analyses of validated opportunities (long-form Markdown).	Free
`get_blog_post`	Full Markdown body of one blog article by slug — built for AI citation.	Free
`query_opportunities`	Filter opportunities by keyword, score, platform, recommendation and category.	Pro
`list_trending_themes`	Pain points trending up right now, sorted by trend change.	Pro
`get_theme`	Theme-level market signal plus the top-scoring opportunities under it.	Pro

Connect

Point your MCP client at the hosted endpoint (recommended) or run the local stdio package. Create a key at Account → API Keys.

{
  "mcpServers": {
    "painspotter": {
      "url": "https://painspotter.ai/mcp/",
      "headers": { "X-API-Key": "psk_live_your_key" }
    }
  }
}

Zero install. Works in Cursor, Claude Desktop, Claude Code and VS Code. Exposes all 7 tools (4 free + 3 Pro).

{
  "mcpServers": {
    "painspotter": {
      "command": "uvx",
      "args": ["painspotter-mcp"],
      "env": {
        "PAINSPOTTER_API_KEY": "psk_live_your_key",
        "PAINSPOTTER_API_BASE": "https://painspotter.ai"
      }
    }
  }
}

The local stdio package exposes the opportunity tools (query_opportunities, get_opportunity). Use the remote endpoint for the full 5-tool set.

Prefer REST?

Every opportunity is also available over a plain REST API — same key, same quota. GET /api/v1/opportunities and GET /api/v1/opportunities/{id}.

Read the REST API reference →

FAQ

What is MCP?

The Model Context Protocol is an open standard that lets LLM clients call external tools. PainSpotter ships an MCP server so models can query its opportunity database directly from your IDE.

Do I need a paid plan?

A free key can call get_overview, get_opportunity, list_blog_posts and get_blog_post (blog tools cost zero quota). Pro unlocks query_opportunities, list_trending_themes and get_theme.

Which clients are supported?

Any MCP-aware client: Claude Desktop, Cursor, Claude Code, VS Code and others. The remote endpoint needs no install.

Where can I find the server source?

The MCP server is open source (MIT) and listed on Glama and Smithery. The package installs from PyPI as painspotter-mcp.