Back to Home
Teamwork icon

mcp

Verified Safe

by Teamwork

Overview

Enables Large Language Models (LLMs) and AI agents to interact with and manage Teamwork.com projects and Teamwork Desk resources via the Model Context Protocol.

Installation

Run Command
TW_MCP_SERVER_ADDRESS=:8080 go run cmd/mcp-http/main.go

Environment Variables

  • TW_MCP_BEARER_TOKEN
  • TW_MCP_API_URL
  • TW_MCP_SERVER_ADDRESS
  • TW_MCP_LOG_LEVEL
  • OPENAI_API_KEY
  • ANTHROPIC_API_KEY
  • GOOGLE_API_KEY

Security Notes

The project demonstrates strong security practices, including a dedicated `SECURITY.md` outlining vulnerability reporting and best practices. It uses standard authentication mechanisms (Bearer tokens, OAuth2), explicitly warns against hardcoding API keys, and redacts authorization headers in logs. The server implements middleware for authentication, request tracing (Datadog APM), and error reporting (Sentry). Input parsing for tool arguments is type-safe, reducing injection risks. A whitelist of MCP methods can bypass authentication for necessary initialization/discovery. The STDIO server can be run in a `-read-only` mode for enhanced safety.

Similar Servers

Stats

Interest Score39
Security Score9
Cost ClassMedium
Avg Tokens1000
Stars11
Forks9
Last Update2025-12-06

Tags

MCPTeamwork.comLLM IntegrationAI AgentsProject ManagementHelpdesk SupportGoAPI Wrapper