Coordinate parallel adversarial agents to review code changes and output a unified triage report.
npx skills add https://github.com/CityBear3/dotfiles --skill review-citybear3
Please help me install this Agent Skill. Skill: review Source: https://github.com/CityBear3/dotfiles/tree/main/claude/skills/review Command: npx skills add https://github.com/CityBear3/dotfiles --skill review-citybear3
It helps engineers catch correctness, robustness, API misuse, performance, and test gaps by running a structured, multi-agent review in parallel and producing a single unified report with triage-ready recommendations.
Use the review skill by asking the AI to run a comprehensive code review for the current branch changes.
High-intent search queries and answers about installing and using this skill.
An adversarial code review runs parallel verification agents and adversarial hunts for API misuse, performance, and test weaknesses. It coordinates 7 agents concurrently to identify defects and output a unified report with triage-ready Must Fix or Should Improve recommendations.
A multi-agent code review coordinates multiple specialized agents in parallel to evaluate changed code. It runs design alignment, scope completeness, test coverage, robustness, API safety, and performance checks concurrently, integrating findings into a single prioritized report.
Perform a context-grounded code review by building a fail-safe context bundle from design documents, plans, project rules, and language hints. This bundle grounds the parallel reviewer agents, ensuring findings align with your specified design constraints and project standards.
Yes, automated code review can triage defects and test coverage gaps by running structured verification checks concurrently with adversarial test-weakness hunts. It integrates these findings into a standardized, action-oriented report with prioritized Must Fix or Should Improve sections.
Yes, multi-agent code review is designed for reviewing files or branches after verification passes or when explicitly requested. It applies parallel agents to identify robustness, correctness, and API safety risks across the changed code.
Automated adversarial code review requires a fail-safe context bundle from design docs and project rules to ground its agents. Without these inputs, the 7 parallel agents may lack the shared context needed to accurately identify scope completeness and design alignment gaps.
skilld-devSynthesize VueUse core v14.3.0 API changes from docs and releases.
nagisanzeninAnalyze source code for architecture conformance, code quality, and performance issues.
mol4Review BMAD code changes with adversarial analysis and triage findings.
chicanoandres702Coordinate multi-agent reviews of software artifacts and generate unified error assessments.
