skills.rest
koseki2580
koseki2580COMMUNITY
Explore all skills by this author →
GitHub

review-reference-creation

Convert pull request review comments into structured YAML review references.

Updated Apr 30, 2026
Search Tags:#code review#deduplication#PR analysis#lesson generation#applicability#YAML output#review references
One-click install
npx skills add https://github.com/koseki2580/skills --skill review-reference-creation
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: review-reference-creation
Source: https://github.com/koseki2580/skills/tree/main/skills/review-reference-creation
Command: npx skills add https://github.com/koseki2580/skills --skill review-reference-creation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps convert one-off pull request review feedback into a reusable, structured review reference so future reviewer agents can apply consistent judgment without copying raw comments.

Core Features & Use Cases

  • Lesson + Applicability Conversion: Converts review intent into a reusable lesson with applicability.applies_when and applicability.does_not_apply_when.
  • Privacy/Safety Guardrails: Strips personal names and emotional phrasing, avoids secrets/tokens/personal/customer data, and keeps code excerpts minimal.
  • Duplicate-Aware Reference Creation: Checks existing review-references/team-lead/references.yaml and merges, variants, or skips creation when appropriate.
  • Backfill Mode Support: Can generate references from merged PRs or historical review comments using structured backfill inputs.

Quick Start

Use the review-reference-creation skill to generate a candidate review reference by providing a PR URL, reviewer role, review comments, code_before/code_after (and optionally code_after), plus a PR summary and any human notes.

Frequently Asked Questions about review-reference-creation

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

FAQPage Schema
How do I turn pull request review comments into reusable rules?▼

You can convert pull request review comments into reusable rules by transforming reviewer intent and code context into a structured YAML review reference. This captures the generalized lesson and applicability conditions for future automated code review workflows.

What is the best way to automate code review deduplication for PR analysis?▼

Automating code review deduplication for PR analysis involves checking existing YAML review references to merge, variant, or skip new entries. This prevents duplicate lessons by comparing newly generated review references against the current repository dataset.

Can I generate review references from merged PRs using historical review comments?▼

Yes, you can generate review references from merged PRs using historical review comments through backfill mode. This processes structured backfill inputs containing historical PR metadata and post-fix code to reconstruct past review lessons.

Does generating YAML review references strip personal data from code excerpts?▼

Yes, generating YAML review references applies safety guardrails that strip personal names, emotional phrasing, and secrets from code excerpts. This ensures the resulting review reference contains only minimal, generalized code context without exposing sensitive data.

What inputs do I need to create a structured review reference from a PR review?▼

Creating a structured review reference from a PR review requires the PR URL, reviewer role, review comments, code before and after changes, and optionally a PR summary. These inputs determine the resolution and extract the generalized lesson for the YAML entry.

Related Skills

Explore More Software Engineering Skills→
Dq666dingerDq666dinger
1

github-code-review

Analyze diffs and PR context to generate structured code review feedback.

Community
Advanced
Patasse97Patasse97

code-reviewer

Automate PR code review with complexity, risk, and SOLID quality checks.

Community
Advanced
strawgatestrawgate

follow-the-pr

Monitor GitHub pull request states for CI changes, reviews, and merge readiness.

Community
Advanced
leifericfleifericf
5

check-correctness

Automate code review for logic errors, boundary conditions, and arithmetic defects.

Community
Intermediate