This skill implements a controlled fix-and-review workflow for automation pipelines. When a review returns a FIX-FIRST verdict (CRITICAL or HIGH gaps), the skill dispatches a separate fixer subagent to apply targeted fixes, then spawns an independent reviewer subagent to validate the changes. The loop repeats once (max 2 fix attempts) and will escalate with a structured report if gaps persist. It enforces separation of concerns (fix vs review), loop limits, and clear escalation messaging.
Use this skill when automated or orchestrated code reviews produce FIX-FIRST results that must be addressed programmatically but require human-readable escalation if auto-fixing fails. Typical triggers: CI review tool returns FIX-FIRST, orchestrator hands off unresolved gaps, or an automated review flags CRITICAL/HIGH issues.
This skill is written for CLI-based agent orchestrators and agent runtimes that support spawn/send semantics (e.g., Genie/CLI agents). It aligns with agents that can run separate subagents and log to task systems.
This skill has not been reviewed by our automated audit pipeline yet.
