Implement a RuleEngineEventHandler that ships in the plugin system introduced by PR #312, generalizing the hard-coded behavior of the MVP's default ErrorHandler (which only pauses workflows on node-error events).
Scope:
- Rule definitions: match on event source/type/severity/message patterns, or combinations.
- Actions: retry workflow step, restart node, run recovery workflow, send notification (delegates to NotificationHandler), pause workflow (matches current ErrorHandler), escalate.
- Configurable via YAML/API.
- Guard rails: max retries, cooldown windows, circuit breakers — prevents infinite recovery loops.
Layers on top of #322 for the action primitives. First substantial scope-cut from #247 Error Manager Service.
Roadmap metadata
Roadmap note: Depends on #338 — rule evaluation against the DB cannot run inline in log_event without slowing ingestion.
Generated as part of the MADSci → 1.0 roadmap. See docs/roadmap.md for context.
Implement a RuleEngineEventHandler that ships in the plugin system introduced by PR #312, generalizing the hard-coded behavior of the MVP's default ErrorHandler (which only pauses workflows on node-error events).
Scope:
Layers on top of #322 for the action primitives. First substantial scope-cut from #247 Error Manager Service.
Roadmap metadata