Communitygithub.com

MigoXLab/webqa-agent

Use WebQA to test websites, web pages, URLs, login flows, search flows, forms, navigation, and core user journeys with an AI browser QA agent.

webqa-agent とは?

webqa-agent is a Claude Code agent skill that use WebQA to test websites, web pages, URLs, login flows, search flows, forms, navigation, and core user journeys with an AI browser QA agent.

対応~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/MigoXLab/webqa-agent/tree/HEAD/skills/webqa

お気に入りのAIに質問する

このエージェントスキルを事前に読み込んだ状態で新しいチャットを開きます。

ドキュメント

WebQA

Use WebQA when the user wants to test a website or page from a real user's perspective.

Route

  • For a normal website, page, URL, feature, login, search, form, navigation, or core-flow test, read references/mini-agent.md.
  • For MCP server installation, API key setup, IDE configuration, or environment variables, read references/setup.md.
  • For authenticated testing with browser cookies, read references/cookies.md before running the test.
  • For API errors, execution timeouts, missing reports, worker-limit errors, or stuck executions, read references/troubleshooting.md.

Defaults

  • If the user provides a URL and no specific objective, run the default mini-agent task: verify the homepage loads, core entry points are visible, and the main interactions show no obvious errors.
  • If the user describes a test objective but gives no URL, ask for the URL before running WebQA.
  • Keep the task focused on one URL and one natural-language objective.
  • Do not treat execution completion as test success; final pass/fail reporting must come from the test report or documented fallback logic.

Boundaries

  • This public skill exposes WebQA's MCP quick-mode testing workflow.
  • Do not promise future capabilities until their reference files exist.
  • Do not load cookies.md, setup.md, or troubleshooting.md during a normal public-page test unless that context is needed.

関連スキル