Review codebases and pull requests for security risks and correctness flaws.
npx skills add https://github.com/GeekKingCloud/skills --skill roast-geekkingcloud
Please help me install this Agent Skill. Skill: roast Source: https://github.com/GeekKingCloud/skills/tree/main/roast Command: npx skills add https://github.com/GeekKingCloud/skills --skill roast-geekkingcloud
This Skill eliminates superficial, inconsistent code reviews that miss critical security risks, correctness flaws, and maintainability debt, delivering structured, evidence-backed feedback instead of vague opinions.
Use the roast skill to review the code in the current workspace and deliver a full security-focused audit report with findings and a grade.
High-intent search queries and answers about installing and using this skill.
A security audit on a pull request identifies critical risks like injection and auth bypasses, delivering severity-rated findings and fix directions before merge. This process prioritizes security flaws over minor style issues to ensure release readiness.
Grading codebase health involves assessing correctness flaws, maintainability debt, and security risks to assign an overall A-F grade. This structured grading aligns with engineering best practices to provide evidence-backed feedback.
Security-first code review for generated code validation targets hidden critical risks and correctness flaws to determine production readiness. It applies strict risk assessments and delivers evidence-backed findings with clear impact and fix directions.
Assessing legacy codebase risks can be automated by performing directed target reviews that identify critical risks, correctness flaws, and maintainability debt. This delivers severity-rated findings and an overall grade without manual inspection.
A strict code review does check for secret exposure and auth bypasses, prioritizing these critical security vulnerabilities over minor style issues. It provides structured, evidence-backed feedback to eliminate superficial security misses.
CarbeneAIAudit codebases for leaked secrets, CI/CD misconfigurations, and vulnerable dependencies.
krzemienskiScore ValidationForge primitives against a baseline and output A-F grades.
vinaykumarvkAssess code across security, quality, UI, and infrastructure domains.
thoreinsteinGenerate prioritized triage checklists for production incidents and log findings to Obsidian.
