
from gascity-packs
A high-rigor, multi-agent code review pipeline using tiered personas and confidence-gated findings for professional software engineering.
This skill implements a sophisticated code review process that spawns parallel sub-agents representing different reviewer personas (Correctness, Security, Performance, etc.). It merges and deduplicates findings into a structured report, providing high-confidence signals on code quality and safety.
Use this before creating a Pull Request, after completing an iterative implementation task, or whenever professional-grade feedback on code changes is required.
Codex, Claude Code, and other advanced agentic harnesses supporting parallel sub-agent spawning and structured JSON communication.
This skill has not been reviewed by our automated audit pipeline yet.