skills.rest
kai-kou
kai-kouCOMMUNITY
Explore all skills by this author →
Website𝕏@k_aik_ouGitHub

pr-review-watcher

Automates self-review, feedback resolution, and auto-merge for GitHub pull requests.

Updated Aug 17, 2026
Search Tags:#automation#code review#pull request#github#auto-merge#self-review#ci monitoring
One-click install
npx skills add https://github.com/kai-kou/gem-hunter --skill pr-review-watcher-kai-kou
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pr-review-watcher
Source: https://github.com/kai-kou/gem-hunter/tree/main/.claude/skills/pr-review-watcher
Command: npx skills add https://github.com/kai-kou/gem-hunter --skill pr-review-watcher-kai-kou

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? After creating a pull request, developers must manually monitor review feedback, respond to comments, resolve threads, verify CI status, and merge. This Skill automates that entire post-PR lifecycle so reviews are handled and merges complete without user intervention. ## Core Features & Use Cases - Layer 1 Self-Review: Launches a project-level code-review skill on every PR, records findings as inline comments, and verifies review completion mechanically before merging. - Automated Feedback Handling: Classifies review findings, commits fixes or documents skip reasons, replies to threads, and resolves them via GitHub MCP or GraphQL. - Status Detection & Session Recovery: Uses check_pending_pr_reviews.py to detect pending PRs (awaiting_review, needs_prompt, needs_response, needs_resolve_check) and resume ownership after session restarts. - Gated Auto-Merge & Deploy: Verifies CI check runs, evidence freshness, and unresolved threads before squash-merging, then routes production deploys through Workers Builds with sprint review gates. - Use Case: After opening a PR, the Skill self-reviews the diff, fixes or answers all findings, waits for green CI, merges automatically, and records a silent summary on the PR thread. ## Quick Start Ask the AI to monitor the pull request you just created, handle all review feedback, and see it through to merge.

Frequently Asked Questions about pr-review-watcher

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

FAQPage Schema
How do I automatically handle pull request review feedback on GitHub?▼

Invoke this Skill after creating a PR. It runs a Layer 1 self-review, classifies each finding as fix or skip, commits corrections, replies to every inline thread, resolves them, and merges once Layer 0 and Layer 1 gates pass.

How to auto-merge a GitHub pull request after CI passes?▼

The Skill verifies quality-checks workflow runs are green, confirms evidence freshness against the head SHA, checks that zero unresolved threads remain, then performs a squash merge via the GitHub MCP merge_pull_request tool without waiting for external reviewers.

Does this work without the gh CLI in cloud environments?▼

Yes. In cloud environments where gh is unavailable, all GitHub operations route through MCP tools such as pull_request_read, resolve_review_thread, and add_reply_to_pull_request_comment. The gh commands in the reference are local-only fallbacks.

What happens if a session times out while watching a PR?▼

On session restart, check_pending_pr_reviews.py detects pending PRs by status (needs_prompt, needs_response, needs_resolve_check) and the Session-Id trailer in the PR body identifies ownership, so the Skill resumes review and merge responsibility automatically.

When does the Skill refuse to merge a pull request?▼

Merging is blocked when CI check runs are red, when the PR body evidence table is stale relative to the head SHA, when unresolved review threads remain, or when Layer 1 self-review verification fails. Bot automation PRs also require an active_session recheck before merging.

Related Skills

Explore More Software Engineering Skills→
saffron-healthsaffron-health
862

push

Commit changes, create or update pull requests, and monitor GitHub checks.

Official
Advanced
l3ochol3ocho

lessons-learned

Capture and search structured sprint lessons with metadata for retrospectives.

Community
Advanced
SigmacodeatSigmacodeat

cross-modal-review

Route work products to an independent AI model for contract-based review.

Community
Advanced
thebtfthebtf

pr-review

Orchestrates autonomous multi-agent PR reviews with parallel AI comment processing and build verification.

Community
Advanced