Mcp-worker
Enables ChatGPT to securely control a single authorized Windows worker through an MCP endpoint, handling job scheduling, approval, and execution with outbound-only connectivity.
README
Mcp-worker
Mcp-worker lets ChatGPT operate one policy-controlled Windows machine through a public MCP server and an outbound-only worker connection.
The repository contains three parts:
src/remote: the Streamable HTTP MCP server and durable relay APIsrc/worker: the outbound worker client and exact-action approval bridgepackages/windows-worker: the complete Windows tool, policy, audit, desktop, clipboard, process, transfer, admin, and .NET helper implementation
The remote server advertises the complete 31-tool catalog even when the worker is offline. Tool execution remains subject to the worker's local policy. Prompt decisions use MCP elicitation and a short-lived ticket bound to the exact request.
What works
- OAuth-protected
/mcpand separately authenticated worker endpoints - SQLite jobs, leases, expiry, deduplicated results, cancellation, and indeterminate mutation handling
- file CRUD and search, process execution and interactive sessions, desktop and clipboard tools
- AppContainer execution through the included .NET 10 helper
- file import and chunked file export through expiring MCP resources
- local SQLite audit records with relay correlation IDs
- current-user scheduled-task scripts with hidden execution, restart behavior, and bounded logs
- Linux relay tests, Windows worker tests, native-helper builds, and Credential Manager smoke verification in CI
Local verification
npm ci
npm run build
npm test
npm run test:windows
npm run native:build
Copy .env.example into the environment of each process before starting it. Run the remote service with npm run remote and the Windows client with npm run worker.
Public deployment still requires choices that cannot be inferred from this repository: domain and DNS control, ingress, administrator support, OAuth registration, the service account, persistent directories, backup and log policy, alerting, and size limits. See docs/owner-setup-checklist.md and docs/operations.md.
Documentation
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.