Your Upwork feed, inside Claude and Cursor.
The UpHunt MCP server connects your AI-scored Upwork job feed to Claude Desktop, Cursor, Cline, and Windsurf. Ask your AI tool to find the best-matched jobs and pull client research in plain language, without leaving your editor.
Works with any MCP client
Live in three minutes
No build step. Point your AI client at the package, paste your API key, and restart.
- 1
Get your API key
Open your UpHunt dashboard and go to Auto-Apply, API & Webhooks. Click Generate API Key and copy it.
- 2
Add UpHunt to your AI client
Drop the config block into your MCP client settings (claude_desktop_config.json, Cursor's mcp.json, etc.) and paste your key.
- 3
Restart and ask away
Restart the client and try: "Find React jobs over $5,000 with a match score above 8 and summarize the top client."
{
"mcpServers": {
"uphunt": {
"command": "npx",
"args": ["-y", "uphunt-mcp"],
"env": {
"UPHUNT_API_KEY": "your_api_key_here"
}
}
}
}What your AI can do
The server exposes read-only tools over your UpHunt data. Research and draft in your AI tool, then apply through the API.
search_jobsSearch your live, AI-scored Upwork job feed by keyword, budget, job type, and minimum match score.
get_jobPull the full details of any Upwork job by URL or ciphertext, including budget, skills, and screening questions.
get_clientRead a buyer's open jobs, hire history, and spend stats before you decide to bid.
list_applied_jobsReview the jobs you've already applied to, with their match scores and current status.
No Upwork automation
The server never logs into Upwork or scrapes the site. It reads from UpHunt's own API, so your Upwork account is never touched.
Your key, your machine
It runs locally and authenticates with your UpHunt API key. Your credentials never pass through a third-party server.
Human stays in control
Reading and drafting happen in your AI tool. Sending proposals is a deliberate step you take through the API or dashboard.
Ready to apply at scale? Use the UpHunt API.
The MCP server is for research and drafting. When you want to auto-apply programmatically, generate AI cover letters, and react to matched jobs over webhooks, the UpHunt API is the full surface, same API key, one credit per successful application.
- Apply to any job by URL or ciphertext
- AI-generated cover letters
- Webhooks for matched jobs
- Typed Node & Python SDKs
Frequently asked
Does the MCP server submit proposals for me?+
No. The MCP server is read-only: it surfaces your scored job feed and job and client data inside your AI tool so you can research and draft. Submitting proposals is handled by the UpHunt API and your auto-apply pipeline, where a human stays in control of what gets sent.
Is it safe for my Upwork account?+
Yes. The server runs locally on your machine and only reads data from UpHunt's own API using your UpHunt API key. It does not log into Upwork, scrape pages, or automate the Upwork web UI, so there is nothing for Upwork to flag.
What do I need to get started?+
An UpHunt account and your UpHunt API key. Generate the key in your dashboard under Auto-Apply, API & Webhooks, drop it into your AI client's MCP config, and restart the client.
Which AI tools are supported?+
Any client that speaks the Model Context Protocol, including Claude Desktop, Cursor, Cline, Windsurf, and Claude Code.
How is this different from the UpHunt API?+
The MCP server is a thin, read-only bridge for AI tools. The UpHunt API is the full programmatic surface, including auto-apply, AI proposal generation, and webhooks. Most teams use both: the MCP to research inside their AI tool, and the API to automate applications.
Put your Upwork feed where you already work.
Connect UpHunt to your AI tool in three minutes and start every search with the best-matched jobs in front of you.