Authenticate with Authorization: Bearer YOUR_TOKEN; Rootly's API follows the JSON:API specification, so request and response bodies use the data.type and data.attributes structure
Fetch reference IDs: GET /v1/severities, GET /v1/teams, and GET /v1/environments to collect IDs needed for incident creation
Create an incident: POST /v1/incidents with data.attributes containing title, severity_id, team_id, and environment_id; the API returns the incident ID and a Slack channel link if Slack is integrated
Update incident fields mid-incident: PATCH /v1/incidents/{id} with updated attributes such as status (mitigated, resolved) or a summary string
Trigger a workflow automation: Rootly workflows fire based on incident lifecycle events (declared, mitigated, resolved); configure triggers in the Rootly UI and use the API to advance incident state to trigger them
Create a postmortem: POST /v1/post_mortems with incident_id; Rootly auto-populates the postmortem template with the incident timeline and signals Jira or Confluence integrations to create linked documents
Known gotchas
Rootly's API follows JSON:API; all filtering uses the filter[field]=value query parameter syntax rather than plain query params — misformatted filters return all records silently
Workflow automations are triggered by incident state transitions and are not directly callable via the API; advancing incident state via PATCH is the mechanism to trigger associated automations
Rootly introduced an AI-agent-first API design in February 2025 with machine-readable Agents JSON descriptors; these enable LLM-based agents to discover and call Rootly workflows but do not change the underlying REST endpoints
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