Project Evaluator
Dispatches autonomous skeptical critic agents to perform deep, multi-dimensional project audits and identify hidden architectural or logic issues.
Project Evaluator is a sophisticated diagnostic tool designed to move beyond simple self-correction by deploying a team of specialized background agents to scrutinize a codebase. By separating the creation process from the evaluation process, it uses a 'skeptical' lens to uncover flaws in architecture, security, user experience, and logic that developers might overlook. Whether focusing on a specific new module or exploring an entire legacy codebase, this skill provides consolidated, severity-ranked reports that ensure high-quality production standards and robust system design.
Características Principales
012 GitHub stars
02Autonomous skeptical critic agents that operate independently from the main chat
03Multi-angle evaluation covering code, security, UX, architecture, and data
04Explore Mode for time-limited, comprehensive project-wide assessments
05Consolidated reporting with deduplicated findings and severity-ranked IDs
06Focus Mode for deep-diving into specific modules like auth or payment flows
Casos de Uso
01Onboarding onto a new codebase by running a deep exploration to find hidden issues
02Reviewing complex feature modules to identify edge cases and architectural debt
03Performing a comprehensive security and logic audit before a major production release