skills.rest
NalinDalal
NalinDalalCOMMUNITY
Explore all skills by this author →
Website𝕏@nalin82929GitHub

code-review

Reviews pull requests with confidence-weighted issue scoring and false-positive suppression.

1|Updated Aug 17, 2026
Search Tags:#code review#checklist#pull request#security#quality gates#false positives#verdict
One-click install
npx skills add https://github.com/NalinDalal/skillset --skill code-review-nalindalal
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/NalinDalal/skillset/tree/main/skills/engineering/code-review
Command: npx skills add https://github.com/NalinDalal/skillset --skill code-review-nalindalal

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code reviews often drown in noise: vague suggestions, style nitpicks, and low-confidence guesses that waste reviewer and author time. This Skill provides a structured review methodology that weights every issue by confidence, suppresses common false positives, and produces a clear APPROVE/WARNING/BLOCK verdict. ## Core Features & Use Cases - Severity-Tiered Checklist: Categorizes findings into CRITICAL (security, data loss, race conditions), HIGH (breaking changes, performance regressions, missing tests), MEDIUM, and LOW so reviewers know what blocks a merge. - Confidence Weighting (0.0-1.0): Every reported issue requires a confidence score, concrete line citations, and a described failure mode; anything below 0.5 without evidence is suppressed. - Verdict & Template System: Outputs a standardized review report with summary, verdict, tiered issues, positives, and a testing checklist. - Use Case: When reviewing a teammate's or AI-generated PR, run the checklist to catch SQL injection or N+1 queries as blockers while filtering out noise like "consider adding error handling" on already-handled paths. ## Quick Start Review the current pull request using the code-review checklist and give me a confidence-weighted verdict with line-cited issues.

Frequently Asked Questions about code-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review a pull request for security vulnerabilities?▼

Check the CRITICAL tier first: look for SQL injection, XSS, CSRF, secrets committed in code, and missing authentication or authorization checks. Any confirmed critical issue blocks the merge with a BLOCK verdict.

How to reduce false positives in AI code reviews?▼

Assign each issue a confidence score from 0.0 to 1.0 and never report anything below 0.5 without an exact line citation and concrete failure mode. Skip vague patterns like "might be slow" without benchmarks or "could be more readable" without specific suggestions.

What should a code review checklist include?▼

Cover four severity tiers: CRITICAL for security and data loss, HIGH for breaking changes, performance regressions, and missing tests, MEDIUM for duplication and long functions, and LOW for naming and style. End with an APPROVE, WARNING, or BLOCK verdict.

When should a code review block a merge?▼

Block when CRITICAL issues exist, such as security vulnerabilities, race conditions, or data loss risks, or when HIGH issues like API contract violations and missing regression tests are present. Medium and low findings should never block a merge.

What should reviewers not comment on during code review?▼

Skip personal style preferences, trivial formatting handled by linters, library choices without security or performance concerns, and variable naming unless truly confusing. Focus on logic correctness, error handling, security, performance, and test coverage.

Related Skills

Explore More Software Engineering Skills→
AvaTar-ArTsAvaTar-ArTs

github-code-review

Review local git diffs and GitHub pull requests with inline comments via gh or REST API.

Community
Advanced
cuenobicuenobi

sonarqube-quality-gate

Analyze code changes against SonarQube rules for quality gate compliance.

Community
Intermediate
yusufcmgyusufcmg

springboot-verification

Automate end-to-end verification of Spring Boot projects with builds, tests, and security scans.

Community
Advanced
VSF-QC-TTSVSF-QC-TTS

common-pentest-methodology

Map attack surfaces and threat models using PTES-aligned penetration testing methodology.

Official
Advanced