review-comment-triage

Classify PR review comments into bounded categories and produce a deterministic execution plan.

4|1|Updated Feb 3, 2026
One-click install
npx skills add https://github.com/danielbaustin/agent-design-language --skill review-comment-triage
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: review-comment-triage
Source: https://github.com/danielbaustin/agent-design-language/tree/main/adl/tools/skills/review-comment-triage
Command: npx skills add https://github.com/danielbaustin/agent-design-language --skill review-comment-triage

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Comment reviews for PRs are noisy and unstructured, making it hard to plan work without losing context. This skill classifies comments into bounded categories and emits a deterministic execution plan that preserves links, file/line context, and uncertainty, enabling safe handoffs without widening scope.

Core Features & Use Cases

  • Classifies comments into actionable_now, already_fixed, stale_or_not_reproducible, follow_on_issue_needed, and blocked_or_operator_decision.
  • Preserves stable identity, file/line context, and review links for traceability.
  • Outputs a structured execution plan with evidence buckets and suggested handoffs for future work.
  • Works with either local payloads or live PR reviews, and enforces non-widening handoffs.

Quick Start

Run the triage against a PR review payload to produce a bounded, evidence-backed execution plan.

Frequently Asked Questions about review-comment-triage

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

FAQPage Schema
How do I create an execution plan from PR review comments?

To create an execution plan from PR review comments, classify each comment into bounded categories like actionable_now or blocked_or_operator_decision. This structured triage process preserves file/line context and review links, yielding a deterministic plan suitable for safe handoffs without widening scope.

What is the best way to triage noisy GitHub PR reviews for handoff?

The best way to triage noisy GitHub PR reviews is to classify comments into evidence-backed buckets such as stale_or_not_reproducible and follow_on_issue_needed. This preserves comment identity and uncertainty, producing a structured artifact that enforces non-widening handoffs for future work.

Can I process local payloads instead of live PR comments for review triage?

Yes, you can process local payloads instead of live PR comments for review triage. The classification mechanism works with either local payloads or live PR reviews, preserving stable identity and context to generate a bounded execution plan for auditing.

How does classifying PR review comments prevent scope widening during handoffs?

Classifying PR review comments prevents scope widening during handoffs by enforcing non-widening boundaries on the output. Comments are sorted into strict categories like already_fixed, preserving uncertainty and context to yield a structured artifact that limits unrequested work.

What categories are used when classifying PR feedback into an execution plan?

Categories used when classifying PR feedback into an execution plan include actionable_now, already_fixed, stale_or_not_reproducible, follow_on_issue_needed, and blocked_or_operator_decision. These bounded buckets capture evidence and suggested handoffs for traceability.

Does review triage work without adding external dependencies to my repository?

Yes, review triage works without adding external dependencies to your repository. The Skill operates independently using internal scripts and references to process PR review payloads and emit a deterministic execution plan.