gh-respond

Read Claude-addressed GitHub comments and write replies to /tmp/comment.txt.

1|Updated Feb 7, 2026
One-click install
npx skills add https://github.com/chriswickett/agentic-github --skill gh-respond
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gh-respond
Source: https://github.com/chriswickett/agentic-github/tree/main/skills/gh-respond
Command: npx skills add https://github.com/chriswickett/agentic-github --skill gh-respond

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The tool provides read-only, automated responses to Claude-addressed comments on GitHub issues and PRs, enabling quick feedback without touching code or triggering changes.

Core Features & Use Cases

  • Read and interpret comments that start with @claude, then generate a clear, helpful reply.
  • Run entirely within a GitHub Actions VM with no code changes to the repository.
  • Useful for issue triage, clarifying questions, and PR feedback where only conversation is needed.

Quick Start

Comment with '@claude' on a PR or issue to trigger a ready-to-respond message.

Frequently Asked Questions about gh-respond

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

FAQPage Schema
How do I automate read-only responses to @claude comments on GitHub pull requests?

To automate read-only responses to @claude comments on GitHub pull requests, trigger the gh-respond Skill within a GitHub Actions VM. It reads the addressed comment, generates a reply, and writes it to /tmp/comment.txt without modifying repository code.

Can I use GitHub Actions for AI-assisted issue triage without granting repository write access?

Yes, you can use GitHub Actions for AI-assisted issue triage without write access by running gh-respond in a read-only mode. It interprets the @claude mention and composes a conversational reply while adhering to strict read-only safety constraints.

What is the best way to trigger an AI assistant for code review feedback on a GitHub issue?

The best way to trigger an AI assistant for code review feedback on a GitHub issue is to comment with @claude. This initiates the read-only response workflow, allowing the assistant to read the issue context and formulate a helpful reply without pushing changes.

How does a GitHub Action process @claude comments to generate automated replies?

A GitHub Action processes @claude comments by running the gh-respond Skill inside the virtual machine. It parses the trigger comment, composes an automated reply based on the context, and outputs the result to a temporary file for read-only feedback.

Are there limitations to using read-only AI assistants for GitHub PR feedback?

Limitations of using read-only AI assistants for GitHub PR feedback include the inability to directly commit code changes or modify repository files. The assistant operates strictly within the GitHub Actions VM to provide conversational triage and clarifying questions.

Does the gh-respond Skill require any external dependencies to answer Claude-triggered comments?

No, the gh-respond Skill does not require external dependencies to answer Claude-triggered comments. It operates entirely within the GitHub Actions VM environment, reading the trigger and composing a reply without needing additional components.