Run blind parallel adversarial code reviews with severity triage and re-review.
npx skills add https://github.com/susomejias/rembric --skill judgment-day-susomejias
Please help me install this Agent Skill. Skill: judgment-day Source: https://github.com/susomejias/rembric/tree/main/.agents/skills/judgment-day Command: npx skills add https://github.com/susomejias/rembric --skill judgment-day-susomejias
Manual code reviews often miss critical issues due to reviewer bias, fatigue, or oversights, leading to bugs and security vulnerabilities making it into production.
Use the judgment-day skill to run a dual adversarial review of the open pull request for the user authentication feature.
High-intent search queries and answers about installing and using this skill.
Blind dual code review runs two independent adversarial reviewers in parallel with identical criteria to catch missed issues. This eliminates reviewer bias and fatigue by classifying findings by severity, ensuring only confirmed fixes get applied.
Run an adversarial pull request audit by launching parallel blind reviewers to evaluate the code. The process enforces a controlled workflow where only confirmed fixes are applied, followed by a mandatory post-fix re-review to validate changes.
Yes, this dual review approach works for architecture slice assessments and new feature implementations. It applies to any codebase evaluation requiring high-confidence issue detection, including high-stakes pull requests and feature audits.
The best way to triage code review findings by severity is to classify them as critical issues, real warnings, theoretical warnings, or suggestions. This severity-based issue triage prioritizes fixes and ensures only confirmed fixes are applied.
The controlled fix workflow prevents unvetted code changes by applying only confirmed fixes and then re-running both reviewers. This mandatory post-fix re-review validates changes before approval, ensuring no unvetted code makes it into production.
Manual code review misses critical security vulnerabilities and bugs due to reviewer bias, fatigue, or oversights. Running parallel adversarial code reviews eliminates these missed issues by enforcing strict blind judging protocols.
dimsourFetch GitHub pull request diffs and run parallel code quality, security, testing, and Iron Law reviewers.
majiayu000Route user requests to registered skill modules and execute them in context.
janewilkinLints Python code with ruff and checks types with mypy.
RoxabiApply code review findings by triaging and fixing them sequentially or interactively.
