Summary
The monitoring you keep meaning to set up, set up while you code.
Scheduling and uptime checks usually get postponed because they mean leaving the editor for a dashboard. Here registering a job, adding a monitor with assertions and reading execution history are all tool calls — and pause and resume exist, so deploys do not have to mean deleting and rebuilding.
What it is
A Go MCP server for Tickstem, exposing cron scheduling, uptime monitoring and heartbeat monitoring as tools over stdio.
What you get
- Cron jobs across their whole life: registered with a name, schedule and endpoint; listed; fetched by id; updated field by field; paused; resumed; deleted
- Execution history for a job, most recent first, for when a schedule stopped firing and you need to know when
- Uptime monitors created with optional response assertions, then listed with status, URL, interval, SSL expiry and the assertions in force
- Monitor checks listed, so a flapping monitor can be read check by check
- Monitors paused and resumed, which is what you want during a planned deploy rather than deleting them
- Heartbeat monitors created and listed, for jobs that report in rather than being polled
Requirements
A Tickstem API key in TICKSTEM_API_KEY. Prebuilt binaries are published for macOS (Apple Silicon and Intel), Linux (Intel and ARM) and Windows, so Go is not required to run it.
Setup effort
One command plus a key — go install github.com/tickstem/mcp/cmd/tsk-mcp@latest, then supply credentials