Feature Request: Official Toggl Track MCP Server
Hi Toggl team,
I’d love to see Toggl offer an official MCP (Model Context Protocol) server for Toggl Track.
MCP is an open standard (created by Anthropic) that lets AI assistants like Claude, ChatGPT, Cursor, and others interact directly with external tools. It’s quickly becoming the standard way AI apps connect to services — think of it as what REST APIs did for web apps, but for AI agents.
Right now, the only options for connecting Toggl to AI assistants via MCP are:
- Pipedream’s hosted server, which only exposes a handful of Toggl actions (can’t even start/stop timers or list projects)
- Several community-built servers on GitHub, which work well locally but aren’t hosted, so they don’t work with web-based AI clients like claude.ai or ChatGPT
An official Toggl MCP server — ideally hosted as a remote server with OAuth or API key — would let users manage timers, view projects, pull reports, and track time directly from any MCP-compatible AI client. The API coverage is already there in Toggl’s v9 API; it just needs an MCP wrapper.
Companies like Slack, GitHub, Atlassian, Linear, and Sentry already offer official MCP servers. For a productivity tool like Toggl, where the whole point is reducing friction around time tracking, this feels like a natural fit.
Would be happy to share more details on what the community has built so far if it helps inform the effort.
Thanks!