Skip to main content
Squasher on-call gives incidents, notifications, agents, and operators one shared responder model. A schedule can be team-wide or project-scoped, with materialized shifts that agents can inspect before routing a handoff.

What on-call includes

  • Native weekly schedules with ordered rotation members.
  • Team-wide schedules, plus project-scoped schedules for service-specific overrides.
  • Current engineer lookup for failed-triage handoff.
  • Shift windows and an iCalendar feed for human calendar clients.
  • Temporary overrides for coverage changes.
  • API, CLI, and MCP Code Mode access for read and confirmed write workflows.

Setup paths

  1. Choose the project, schedule name, timezone, handoff day, and handoff time. 2. Add members in rotation order. 3. Decide whether the schedule is team-wide or project-scoped. 4. Verify the current engineer and upcoming shift window before relying on the schedule.

CLI

Write commands should be run only after the user confirms the intended schedule or override.

API and MCP

Use the On-call API reference for schemas and operation IDs. The required API key scopes are on_call:read for list/current/member/shift operations and on_call:write for create, enable/disable, and override operations. The Squasher MCP server stays Code Mode only. Search for onCall operations, inspect the schema returned by search, then call exposed operations with execute. There are no dedicated MCP on-call tools.

Migration guides

Agent handoff

Use this prompt from the dashboard or paste it into an assistant that has Squasher CLI, API, or MCP access: