ratchet:watch

Monitor Ratchet PRs for merge conflicts, CI failures, and review comments.

1|Updated Mar 12, 2026
One-click install
npx skills add https://github.com/netbrain/ratchet --skill ratchet-watch
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ratchet:watch
Source: https://github.com/netbrain/ratchet/tree/main/skills/watch
Command: npx skills add https://github.com/netbrain/ratchet --skill ratchet-watch

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ratchet:watch eliminates the need for manual PR triage by continuously checking active Ratchet PRs for merge conflicts, CI failures, and actionable review comments and surfacing them as structured discoveries in .ratchet/plan.yaml so work is prioritized and tracked.

Core Features & Use Cases

  • Watches PRs referenced in .ratchet/plan.yaml and polls on a configurable interval.
  • Detects merge conflicts and CI failures and creates deduplicated discovery entries tied to the originating issue and milestone.
  • Fetches review comments and review verdicts, classifies them (actionable vs informational), records feedback as discoveries, and optionally spawns response agents unless --no-respond is set.
  • Maintains a persistent .ratchet/watch-state.json to avoid reprocessing and handles authentication checks and malformed-state recovery.
  • Use case: Run during ratchet:run to surface blocking CI failures and reviewer change requests automatically so response agents or teams can act immediately.

Quick Start

Run /ratchet:watch from your project workspace to start monitoring Ratchet PRs and automatically create discoveries for conflicts, CI failures, and actionable review comments.

Frequently Asked Questions about ratchet:watch

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

FAQPage Schema
How do I automatically monitor GitHub pull requests for CI failures and merge conflicts?

You can continuously monitor GitHub pull requests by polling active PRs to detect CI failures, merge conflicts, and actionable review comments. Structured discoveries are generated and tracked automatically so work can be prioritized without manual triage.

What is the best way to automate PR triage for blocking code review feedback?

Automating PR triage involves fetching review comments and verdicts, classifying them as actionable or informational, and recording the feedback as discoveries. Response agents can optionally be spawned to address the review feedback immediately unless explicitly disabled.

Do I need GitHub CLI authentication to monitor pull request statuses?

Yes, authenticated GitHub CLI access is required to monitor pull request statuses. Additionally, the repository needs read/write permissions to update the plan configuration and a writable state file is necessary for deduplication during periodic polling.

How does PR monitoring handle duplicate issues during periodic polling?

PR monitoring handles duplicate issues by maintaining a persistent state file that tracks processed entries. This deduplication mechanism ensures that existing merge conflicts, CI failures, and review comments are not reprocessed or duplicated across polling intervals.

Can I stop automated response agents from replying to PR review comments?

Yes, you can stop automated response agents from replying to review comments by setting the no-respond flag. This restricts the monitoring process to only surfacing discoveries without spawning any agents for feedback.

Why is my PR monitoring failing to recover from a malformed state file?

PR monitoring state file recovery depends on the persistent watch-state configuration. If the state file becomes malformed, the monitoring system includes recovery mechanisms to handle the malformed state and resume polling without losing tracked PR data.