respond-reviewer

What problem does it solve?

Manually responding to peer reviewer comments is time-consuming and error-prone, especially for letters with dozens of comments spanning different types (prose edits, new analysis, citation gaps, structural framing changes). This Skill eliminates that manual work by automating comment classification, specialist routing, and rebuttal letter assembly.

Core Features & Use Cases

• Structured Comment Parsing: Auto-detects Markdown or LaTeX reviewer letters and parses them into stable, uniquely ID'd comments for consistent tracking across re-runs.
• Supervisor-Led Classification: Dispatches a supervisor agent to label each comment as prose, analysis, citation, clarification, or structural, with an ethical guardrail that routes structural (framing/scope) comments to human review instead of auto-dispatch.
• Specialist Agent Dispatch: Routes non-structural comments to the correct specialist agent (paper-writer for prose/clarification, analysis-implementer for analysis, literature-curator for citations) to draft accurate, context-aware responses.
• Rebuttal Letter Assembly: Assembles all responses into a formatted LaTeX or Markdown rebuttal letter, evaluates response quality via a second supervisor pass, and writes all run state to durable files for audit and re-runs.
• Use Case: A researcher submitting to a top-tier conference receives a 12-comment review letter. They run this Skill to generate a complete, formatted rebuttal letter in minutes, with only 2 structural comments requiring their manual input.

Quick Start

Invoke the respond-reviewer skill with the file path to your peer review letter to automatically generate a complete, formatted rebuttal letter with classified, agent-drafted responses to each comment.