autonomy rollback execute

Execute a rollback action against the target subsystem

Synopsis

Dispatch a rollback action to the appropriate subsystem. The action is
attributed to the AUTONOMY_OPERATOR identity and emits an audit record.

Run 'rollback preview --target <kind>' first to review safety class and known
limitations before executing.

Resource flag usage:
  rollout_plan / rollout_stage  --resource <plan_id>
  ha_leader_resign              --resource current

Strategy flag usage:
  rollout_plan    --strategy retry|rollback
  rollout_stage   --strategy skip_failed
  ha_leader_resign  (no --strategy required)

relay_deadletter is not orchestrated here; use edgectl deadletter retry|purge.

Usage

autonomy rollback execute [flags]

Options

      --orchestrator-url string   orchestrator base URL (env: AUTONOMY_ORCHESTRATOR_URL)
      --output string             output format: "text" or "json" (default "text")
      --reason string             operator-supplied reason for the rollback action (recorded in audit trail)
      --resource string           resource ID: plan_id for rollout targets; 'current' for ha_leader_resign
      --strategy string           recovery strategy: rollout_plan → retry|rollback; rollout_stage → skip_failed
      --target string             rollback target: rollout_plan, rollout_stage, ha_leader_resign

See also

• autonomy rollback — Unified rollback orchestrator for fleet, HA, and relay targets