Error Recovery: Recovery workflow framework + retry/cooldown/circuit-breaker primitives

[RyanTheRobothead]

Workflow-engine primitives that the RuleEngineEventHandler invokes as 'actions':

• Retry semantics (per-step, per-workflow, with backoff)
• Cooldown windows (don't retry the same failure within N seconds)
• Circuit breakers (open after N failures, half-open after cooldown, close after success)
• Recovery-workflow execution (run a named workflow as a reaction to an event)

These hook into the existing workflow engine and are consumed by the EventHandler plugins from PR #312.

---

Roadmap metadata

• Phase: 1.0-rc — Hardening
• Project: Advanced Error Recovery
• Blocked by: Feature error handling MVP #312, Chore: Docker Healthchecks for Managers and Nodes #239
• Blocks: Error Recovery: Re-decide Event-Manager-plugins vs dedicated Error Manager #325

Generated as part of the MADSci → 1.0 roadmap. See docs/roadmap.md for context.