lightweight-review

Reviews code diffs and pull requests for defects, regressions, and missing tests.

Updated Feb 26, 2025
One-click install
npx skills add https://github.com/adschill/dev-config --skill lightweight-review-adschill
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: lightweight-review
Source: https://github.com/adschill/dev-config/tree/main/skills/lightweight-review
Command: npx skills add https://github.com/adschill/dev-config --skill lightweight-review-adschill

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code changes often ship with hidden defects, regressions, or insufficient test coverage because full QA reviews are too heavy for everyday diffs. This Skill provides a focused, high-signal review workflow that surfaces correctness, security, and maintainability risks quickly. ## Core Features & Use Cases - Severity-Ranked Findings: Reports issues in strict severity order (critical, high, medium, low) with higher-confidence findings listed first. - Structured Review Workflow: Defines scope, identifies behavior changes and invariants, checks edge cases and failure paths, and evaluates test sufficiency. - Standardized Output Format: Uses a shared review reporting format with compact areas like Correctness, Security, Maintainability, and Test Coverage. - Use Case: A developer opens a pull request and asks for a focused review; the Skill scans the diff, flags a missing null check as high severity, and notes that the new error-handling path lacks test coverage. ## Quick Start Ask the assistant to use lightweight-review to review this diff for correctness risks and missing tests.

Frequently Asked Questions about lightweight-review

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

FAQPage Schema
How do I get a quick code review of a pull request diff?▼

Ask for a lightweight review of the diff or patch. The review defines scope first, identifies behavior changes and invariants, then reports findings in strict severity order from critical to low with actionable recommendations.

What severity levels are used in code review findings?▼

Findings use four labels: critical for outages, data loss, or security breaks; high for major bugs with broad impact; medium for bounded correctness or maintainability risks; and low for minor issues or clarity gaps.

When should I use a lightweight review instead of a full QA review?▼

Use lightweight review for focused feedback on diffs, patches, or pull requests. It is not intended for full pre-merge QA, broad validation requests, pure feature implementation, or debugging sessions without a review deliverable.

Does lightweight review check test coverage for changed code?▼

Yes, evaluating whether tests are sufficient for the changed behavior is an explicit step in the workflow. Gaps in test coverage are reported as findings under areas like Test Coverage.

What output format does the code review produce?▼

The review follows a shared reporting format with compact areas such as Correctness, Security, Maintainability, and Test Coverage. Findings are ordered by severity, with higher-confidence items listed first within each level.