What is the Nostr event kind used?
The MCP server sends a kind 5109 event to request visual computer interaction help and listens for kind 7000 events with invoices and prices from humans.
@dtdannen
Basic information
MCP Server to ask humans for help
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"unstuckmcp": {
"command": "python3",
"args": [
"-m",
"venv",
"venv"
]
}
}
}Tool list not available yet
This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.
We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.
unstuck-ai is an MCP server that enables AI agents to instantly pay humans sats (Bitcoin) to solve visual roadblocks such as captchas, web navigation, and computer use tasks. It operates via a Nostr marketplace and includes both the MCP server and a web app where humans can bid on tasks, complete them, and receive payment.
Set up the MCP server by installing dependencies (cd mcp_server && pip install -r requirements.txt), configuring environment variables in mcp_server/.env (Digital Ocean Spaces credentials and a Nostr Wallet Connect string), then start the server with fastmcp run unstuck_ai/server.py:mcp --transport sse. In another terminal, run goose session --with-remote-extension http://127.0.0.1:8000/sse and prompt Goose to use the unstuck-ai tool for visual help.
The MCP server sends a kind 5109 event to request visual computer interaction help and listens for kind 7000 events with invoices and prices from humans.
You need a Nostr private key (for the AI agent), a Nostr Wallet Connect string (to pay invoices), and Digital Ocean Spaces credentials (or any boto3 provider for screenshot storage).
The README provides instructions specifically for Block's Goose. It is an MCP server, so it may work with other MCP‑compatible clients, but only Goose setup is documented.
No. The README lists many development items marked as incomplete, including executing the human's action automatically (clicking, dragging) and moving screenshot hosting to a Blossom server.
A human selects a job on the web app, generates a Lightning invoice for the amount, broadcasts a kind 7000 event, and once the AI agent pays the invoice, the human sends the result.

Handl speaks the Model Context Protocol. Your whole billing operation — projects, invoices, clients, reports, approvals — as tools any AI assistant can call, inside your autonomy rules.

Request dedicated senior developers to work on your project, write job descriptions, and prep technical interviews — all without leaving your Claude chat. Just describe what you need, and Lemon.io MCP does the rest. 1–3
A curated list of references for MLOps

Yuppsy MCP: build and publish real estate websites from your AI assistant. Yuppsy lets your AI assistant create live websites for real estate agents and individual property listings. You don't need a designer or any code

DeepMark helps teachers deliver rapid, consistent marking with meaningful feedback for every student — in a fraction of the time. What once took a week, now takes one free period.
Comments