CleanJobData

Jobs MCP Server

Search and read live job listings directly from Claude, Cursor, or any other Model Context Protocol client — no code, no scraping, just a job search MCP server your AI assistant can call directly.

Works on every plan
The MCP server is just a thin wrapper around this same API, authenticated with your regular API key. Every plan can use it. On plans where application_url isn't included directly in search results (e.g. the Solo MCP plan), the server exposes a get_apply_link tool to fetch it — see Unlock Application Link.

1. Get an API key

Generate one from your dashboard if you haven't already.

2. Connect your client

Option A works with any MCP client (Claude Desktop, Cursor, Claude Code, others). Option B is a one-line shortcut, but only if you're using Claude Code specifically. Both point at the same hosted server — neither requires installing anything locally.

Option A: paste a config file

Paste this into claude_desktop_config.json (Claude Desktop), your Cursor MCP settings, or a project's .mcp.json (Claude Code), replacing the key below with your own.

claude_desktop_config.json / .mcp.json
{
  "mcpServers": {
    "cleanjobdata": {
      "url": "https://mcp.cleanjobdata.com/mcp",
      "headers": { "Authorization": "Bearer YOUR_CLEANJOBDATA_API_KEY" }
    }
  }
}

Option B: Claude Code CLI

If you're using Claude Code, this one command registers the server — no config file to edit:

Terminal
claude mcp add cleanjobdata --type http https://mcp.cleanjobdata.com/mcp \
  --header "Authorization: Bearer YOUR_CLEANJOBDATA_API_KEY" \
  --scope user

--scope user makes it available across all your projects, not just the one you run this in. Use /mcp inside a Claude Code session to confirm it's connected.

3. Restart your client

Once restarted, tools like search_jobs, get_job, and search_companies are available to ask about directly in your conversation.

Available tools

  • search_jobs — search by title, location, remote, salary, and more
  • get_job — full detail for a specific job, including description
  • get_apply_link — unlock the real application URL for a job
  • list_expired_jobs — ids of jobs that recently expired, for keeping a local sync in step
  • search_companies — search companies by name or website domain
  • get_company — company detail, including enrichment data
  • suggest_locations — autocomplete city/state/country for geo filters