multi-agent-review
Multi-agent review orchestration dispatching parallel specialist reviewers with line-level evidence collection.
SKILL.mdskills/multi-agent-review/SKILL.md
GitHubMulti-Agent Review Skill
Review Lanes
- Reviewer A (Correctness): Logic defects, state leaks, unhandled exceptions, race conditions.
- Reviewer B (Security): Auth, secrets, injection, privilege escalation, OWASP Top 10.
- Reviewer C (Architecture): Separation of concerns, modularity, dependency boundaries.
- Reviewer D (Adversarial): Hostile payload tracing, edge cases, bypassed validations.
- Reviewer E (Tests): Test coverage, boundary value analysis, regression protection.