resolve-checks

Diagnose and fix failing CI checks and PR review feedback on GitHub pull requests.

1.7k|91|Updated Feb 13, 2025
One-click install
npx skills add https://github.com/flowglad/flowglad --skill resolve-checks
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: resolve-checks
Source: https://github.com/flowglad/flowglad/tree/main/.claude/skills/resolve-checks
Command: npx skills add https://github.com/flowglad/flowglad --skill resolve-checks

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Failing CI checks and unresolved PR review comments block merges and waste time on slow CI round-trips. This Skill runs the full test suite locally first, diagnoses failures, fixes flaky tests at the root cause, and systematically addresses reviewer feedback until the PR is green.

Core Features & Use Cases

  • Local-First Test Execution: Runs type checking, linting, unit, database, RLS, frontend, integration, and behavior tests sequentially before checking CI status.
  • Flaky Test Diagnosis: Identifies non-deterministic data, race conditions, timing dependencies, and shared-state issues, then fixes the underlying cause instead of re-running CI.
  • PR Review Feedback Triage: Categorizes each review comment as actionable, intentional, already addressed, or invalid, then implements fixes or replies with rationale.
  • Use Case: After pushing a branch, CI shows a failing RLS test and a reviewer requested changes. Use this Skill to reproduce the failure locally, fix the policy, address the comments, push, and verify all checks pass.

Quick Start

Resolve all failing CI checks and address review feedback on my current branch's pull request.

Frequently Asked Questions about resolve-checks

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

FAQPage Schema
How do I fix failing CI checks on a GitHub pull request?

Run the full test suite locally first to reproduce failures faster than CI round-trips. Fix each failure, push the changes, then verify the PR status checks turn green using the GitHub API.

How do I diagnose and fix flaky tests that pass locally but fail in CI?

Run the failing test 10 or more times locally to identify intermittent behavior. Common root causes include non-deterministic data, race conditions, timing dependencies, and shared state between tests; fix the underlying issue rather than re-running CI.

How should I handle PR review comments before merging?

Categorize each comment as valid and actionable, intentional design, already addressed, or invalid. Implement fixes for valid feedback, reply with rationale for intentional choices, and resolve addressed threads before re-requesting review.

Why do RLS tests fail when run in parallel?

Row Level Security tests modify database roles and shared state, so parallel execution causes flaky results. They must run serially with a maximum concurrency of one to ensure isolation.

What test types does this workflow cover?

The workflow covers type checking and linting, unit tests, database tests, frontend React tests, RLS policy tests, integration tests with real APIs, and behavior tests across dependency combinations, all run via Bun.