Developers

API & MCP reference

Query your status, monitors, incidents and uptime programmatically — or point any MCP-compatible AI client (Claude, Cursor) straight at your data. Create a token under Admin → API & MCP.

Authentication

All programmatic calls use a bearer token. Generate one under Admin → API & MCP, choosing the abilities it needs: read, write, or admin. Tokens are shown once and stored as a SHA-256 hash — keep them secret.

HTTP header

Authorization: Bearer hst_your_token_here

Base URL: https://status.yt

MCP endpoint

A single JSON-RPC 2.0 endpoint implementing the Model Context Protocol. Point Claude Desktop, Cursor, or any MCP client at it with a read token and ask questions like “is prod down?” or “30-day uptime of the API?”. Streamable HTTP (SSE) is supported via Accept: text/event-stream.

POST https://status.yt/api/mcp

List available tools (curl)

curl -X POST https://status.yt/api/mcp \
  -H "Authorization: Bearer hst_your_token" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Call a tool

curl -X POST https://status.yt/api/mcp \
  -H "Authorization: Bearer hst_your_token" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call",
       "params":{"name":"get_status_summary","arguments":{}}}'

Claude Desktop config (claude_desktop_config.json)

{
  "mcpServers": {
    "statusyt": {
      "url": "https://status.yt/api/mcp",
      "headers": { "Authorization": "Bearer hst_your_token" }
    }
  }
}

MCP tools

Tool	Ability	Description
`get_status_summary`	read	Overall status of your status pages and their components.
`list_status_pages`	read	List your status pages.
`list_monitors`	read	List monitors with their latest status and latency.
`list_incidents`	read	List incidents (pass only_open=true for active ones).
`get_uptime`	read	Per-component uptime % over the last N days (default 30).
`create_incident`	write	Open a new incident on a status page.
`add_incident_update`	write	Post an update to an existing incident by public_id.
`resolve_incident`	write	Resolve an incident with a final message.

Write tools require a token with the write ability.

Heartbeat monitors

For cron jobs and background workers, create a Heartbeat (push) monitor and have your job ping its URL on each successful run. Miss the grace window and the monitor goes down. No auth token needed — the URL contains a secret token.

Ping from a cron job

curl -fsS -m 10 https://status.yt/heartbeats/YOUR_MONITOR_TOKEN >/dev/null

# Example crontab entry:
*/5 * * * * /path/to/job.sh && curl -fsS https://status.yt/heartbeats/YOUR_MONITOR_TOKEN

Status badges

Embed a live SVG badge in a README or site. Available for public status pages.

Markdown

![status](https://status.yt/status/YOUR_SLUG/badge.svg)
![uptime](https://status.yt/status/YOUR_SLUG/uptime-badge.svg)

Health check

Unauthenticated liveness endpoint for your own monitoring.

GET https://status.yt/api/health

Response

{ "ok": true, "time": "2026-07-18T12:00:00+00:00" }

Ready to build?

Create a token and start querying in under a minute.

Start free