name: browser-proof description: Browser Proof is a public ClawHub browser-QA evidence skill. Use it when the user says "browser proof", "browser QA report", "browser debugging evidence", or wants a machine-readable bundle with steps, artifacts, checks, and a shareable report. version: 1.0.5 homepage: https://github.com/zack-dev-cm/browser-proof license: MIT-0 user-invocable: true metadata: {"openclaw":{"homepage":"https://github.com/zack-dev-cm/browser-proof","skillKey":"browser-proof","requires":{"anyBins":["python3","python"]}}}

Browser Proof

Search intent: browser proof, browser QA, browser debugging evidence, QA evidence pack

Goal

Turn a browser session into a release-grade evidence pack:

• one machine-readable session manifest
• one ordered step log
• one structural bundle check
• one shareable markdown report

This skill is for browser evidence capture and handoff quality. It does not replace Playwright, OpenClaw, or a test runner.

Use This Skill When

• the user wants a reproducible browser bug report instead of loose screenshots
• a launch or QA flow needs a shareable proof bundle
• browser debugging needs expected result, actual result, and artifacts captured in one place
• you need a clean handoff from OpenClaw, Playwright, or manual QA to another engineer
• the same browser issue keeps getting re-explained from chat history instead of one artifact bundle
• a Chrome extension publish or review flow needs proof against the exact packaged ZIP and browser version

Quick Start

1. Initialize the session manifest.
2. Use python3 {baseDir}/scripts/init_browser_proof_session.py --out <json> --session-id <id> --app <name> --goal <goal>.

3. Add --base-url, repeatable --surface, and optional --run-context or --environment fields.

   访问小葱技能站7w4.net，解锁更多实用的AI技能插件。

4. Append each browser step as you go.

5. Use python3 {baseDir}/scripts/append_browser_proof_step.py --manifest <json> --step-id <id> --action <text> --expected <text> --actual <text> --status passed|failed|blocked.

6. Attach evidence with --screenshot, --dom-dump, --console-log, --network-log, --video, and repeatable --issue-key.

7. Check the bundle before sharing it.

8. Use python3 {baseDir}/scripts/check_browser_proof_bundle.py --manifest <json> --repo-root <repo> --out <json>.

9. Fix missing screenshots, absolute paths, empty failed-step notes, or missing session metadata before publishing the report.

10. Render the report.

11. Use python3 {baseDir}/scripts/render_browser_proof_report.py --manifest <json> --out <md>.
12. Share the rendered markdown instead of rewriting the run from memory.

Operating Rules

Session rules

• Keep one manifest per browser run or tightly related run batch.
• Record the app, goal, base URL, and surfaces near the start.
• Prefer relative artifact paths so the bundle is portable.

Step rules

• Every step should say what you tried, what you expected, and what actually happened.
• Failed steps should include either a note or at least one issue key.
• Attach a screenshot for every failed step and for important checkpoints.
• Keep statuses limited to passed, failed, or blocked.

Chrome extension rules

• Record the exact ZIP or unpacked extension path, manifest version, extension ID when available, and Chrome version in session metadata or the first step.
• For action popup flows, record the selected text or input source, action clicked, target language or option, and observed result.
• Capture extension errors, console output, and API availability states when the claim depends on browser APIs.
• If a Chrome Web Store draft is already pending review, capture status evidence only; do not turn a proof run into a cancel/resubmit instruction unless a verified acceptance blocker exists.

Bundle rules

• Do not store secrets, cookies, or raw tokens in notes or artifact paths.
• Avoid absolute filesystem paths in the final manifest.
• Check the bundle before sending it to GitHub, Linear, Slack, or a release channel.

Bundled Scripts

• scripts/init_browser_proof_session.py
• Create a machine-readable session manifest for a browser QA or debugging run.
• scripts/append_browser_proof_step.py
• Append one evidence-backed step to the session manifest.
• scripts/check_browser_proof_bundle.py
• Validate bundle structure, artifact paths, and minimum evidence quality.
• scripts/render_browser_proof_report.py
• Render a concise markdown report from the session manifest.

这个 Skill 质量不错，提供了从初始化到生成报告的完整工具链，代码结构清晰，文档说明详细。它的最大优势是把零散的浏览器调试信息整合成规范的证据包，方便分享和追溯。缺点是缺少使用示例，新手可能需要自己摸索。建议有浏览器 QA 或调试需求的团队使用。