사용 방법
- 설치 명령을 복사하고 자신의 에이전트 환경에 설치하기 전에 소스 패키지를 확인하세요.
- 필요한 도구, 인증 정보, 지원 환경은 원본 문서에서 확인하세요.
- 입력과 원하는 산출물을 설명하고 출처와 결과를 대조하세요.
npx skills add https://github.com/firecrawl/firecrawl-workflows --skill firecrawl-qa시작 전 확인
- 설치에는 모델 구독, API 한도, 파일 접근 권한이 포함되지 않습니다.
- 별점 수와 설치 수는 활동 지표이며 독립적인 품질·보안 검증이 아닙니다.
원본 Skill 문서
원본 저장소의 제목, 예시, 코드, 표, 링크, 이미지를 유지해 표시합니다.
Firecrawl QA
Use this to test a live site and return a unified QA report.
Onboarding Interview
Infer the URL, QA focus, and output format from context. If the target URL is clear, proceed immediately.
Ask at most 1-3 concise questions only if blocked, such as the URL, the focus area, or credentials/constraints for protected flows.
Firecrawl Collection Plan
Use Firecrawl map to discover pages. Use Firecrawl browser for interactions, forms, navigation, and responsive/manual checks when available. Use scrape for page content and link extraction.
Parallel Work
If appropriate, use sub-agents or equivalent parallel task runners:
- Full: Navigation and Links, Forms and Interactions, Content and Visual, Error States.
- Forms: Form Discovery, Happy Path, Edge Cases, Validation.
- Navigation: Sitemap, Nav Testing, Link Checker, Routing.
- Responsive: Desktop, Tablet, Mobile, Interaction.
- Performance: Page Load, Asset Audit, Content Efficiency, Comparison.
Each tester should return severity, URL, description, evidence, and reproduction steps.
Final Deliverable
markdown
# QA Report: [Site]
## Summary
- Health score: [x/10]
- Pages tested: [count]
- Issues found: [critical/major/minor]
## Critical Issues
[C-1] URL | Description | Steps to reproduce | Expected vs actual
## Major Issues
[M-1] URL | Description | Steps to reproduce
## Minor Issues
[m-1] URL | Description
## Positive Observations
[What works well]
## Pages Tested
[URLs]
## Agent/Test Summary
[Who tested what]
## Rerun Inputs
workflow: firecrawl-qa
url: [url]
focus: [full/forms/navigation/responsive/performance]Quality Bar
- Include reproduction steps for functional issues.
- Do not report speculative bugs without evidence.
- Deduplicate findings across testers.
