Authenticate to the Grafana Cloud Synthetic Monitoring API using a Synthetic Monitoring token obtained from the SM plugin UI; pass it as Authorization: Bearer YOUR_TOKEN in all requests
POST to /api/v1/check to create a new check; specify job (name), target (URL or host), frequency (milliseconds between executions), timeout, probeIDs (probe location IDs), and settings (http, ping, dns, or tcp)
For an HTTP check populate the settings.http object with method, headers, and body; add validations such as failIfBodyMatchesRegexp or failIfStatusNot2xx to assert content correctness
GET /api/v1/probe to list available probe locations and their IDs; choose probes from multiple regions to get geographic coverage
Retrieve check results and uptime metrics from the Prometheus-compatible metrics endpoint exposed by the Synthetic Monitoring datasource in Grafana Cloud
Add alerting on the probe_success metric (value 0 means failure) using Grafana unified alerting rules targeting the SM datasource
Known gotchas
Probe frequency and timeout are specified in milliseconds in the API but displayed in seconds in the UI; a mismatch when scripting from UI values is a common source of validation errors
The Synthetic Monitoring API uses a separate token type from the standard Grafana Cloud API key; using a general-purpose API key returns 401 errors without a clear explanation
Checks created via the API do not automatically appear in a dashboard; you must build or provision a dashboard that queries the probe_success and probe_duration_seconds metrics from the SM data source
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp