vkirill/claude-lane-stack/browser-qa
Claude Code and agent skill by vkirill/claude-lane-stack
- community
Checking that the SKILL.md file is still in the repository.
Live browser QA + Playwright-compatible replay. Disk .agents/qa/. Use when (terminal Lane Stack): проверь в браузере, кликни сценарий, вёрстка 375, screenshot QA. SKIP: Lane Pilot chats (PM calls `lane_pilot_browser_qa`); DESIGN.md taste (design-lead); product Vue/CSS (writer lane); L0/L1 unit/API (lane-contract).
Install
npx vskill@latest install vkirill/claude-lane-stack/browser-qaSKILL.md
The same SKILL.md was published earlier as vkirill/claude-lane-stack/browser-qa.
Browser QA
Adapted from lee-to/ai-factory#154 replay rules. Paths and gates are lane-stack. No .ai-factory, no human «работает или нет».
Inside a Lane Pilot chat (LANE_PILOT_AGENT_TYPEset / toolslane_pilot_*present) do not run this file's replay and digest procedure. The PM checks an accepted task withlane_pilot_browser_qa; the check runs in a child thread throughbb browser-automation(no MCP browser) and ends with the one fenced JSON verdict its prompt asks for, changing no files..agents/qa/replay,qa-digest,chrome-qa,browser-qa-jevandchrome-devtoolsbelow are the terminal harness. The PM does not click in its own chat.
Info
If $ARGUMENTS is exactly info, print references/info.md verbatim (Russian), then stop.
Layout
Copy structure from references/ next to this skill. English on disk. Chat may be Russian. Redact tokens, cookies, OTP, password, jwt → [REDACTED].
Cases
PM / SPEC acceptance → cases.md. One ## TC-00N heading per case (see references/cases.md). Fields: Surface, Priority, Viewport, Preconditions, Steps, Expected.
| Surface | This agent | |
|---|---|---|
browser-ui | Required. Live browser. Replay first. | |
cli / backend-test / api | Do not click. Note blocked: use L0/L1. No human array-checking. | |
human | Leave pending. The PM asks the operator. |
Default viewports if omitted: 375 768 1280.
Bindings
tree is clean if git status --porcelain has nothing except .agents/qa/; else dirty.
target_fingerprint input is QA TARGET\n<class>\n<normalized_base_url>\n. Normalize: drop userinfo, query, fragment; keep scheme/host/port/base path; strip trailing slash except origin /. Class: local | staging | preview | production | unknown.
Matching replay = embedded case_digest + embedded target_fingerprint + file script_digest all equal REPORT's last proven values.
Stale (uncheck, keep old evidence, do not count as current):
- HEAD or
treechanged → all results stale; scripts stay reusable if matching cases.mdsource_digestchanged → stale only cases whosecase_digestchanged- new TC → pending; removed TC → Stale / removed
- script bytes ≠ recorded script digest → that case stale, needs proof run
- missing digest metadata → unproven, not matching
Replay (PR #154)
One case → replay/TC-00N.js. Smallest async (page) => { … }. No imports, no new npm deps. Navigation uses the literal "BASE_URL" placeholder. At run time substitute normalized_base_url in memory; do not rewrite the file just to inject the host.
- Write the script before the first click. That first execution is the proof. Do not auto-run a new/updated script a second time.
- If mechanics were clarified after the observed run and the file changed, mark
unproven. Repeat only after preconditions restored. Stateful (submit, pay, delete, email, permissions) needs a fresh go-ahead that says this is a repeat. - Record
script_digestas proven only after that exact file executed. Until then:n/a, proofunproven. - Before repair/rebind:
cptoreplay/history/TC-00N-<old_script_digest>.js. Keep the first error in REPORT. Never run history. - Missing selector = product fail until the same control and behavior are proven in-browser and only the locator changed. Do not pick a similar element. Syntax/runner errors may be repaired (new proof). Assertion mismatch = fail, not a reason to regenerate the script.
- Target mismatch → stop. Rebind changes
script_digestand needs a new proof. Never treat a staging script as matching localhost. - After HEAD/
treechange: replay every matching script (old pass and fail). Capability and side-effect rules still apply. provenmeans the script ran, not that the product passed.- Discrete chrome-devtools actions: treat the file as the program; do not rewrite steps because the surface is MCP not
npx playwright. - After replay, at most one bounded exploratory pass when the change is cross-cutting (nav, auth, layout, shared state) or a replay needed repair. Skip when the fix is narrow and matching replay already covers it. New bug → REPORT only; add
## TC-00Nfirst, then a script.
Delegation (stages.browser_qa)
| provider | Who clicks | Command | |||
|---|---|---|---|---|---|
jev (default) | CDP snapshot → Jev Choice (click/type/done). backend = chrome-qa \ | live-chrome \ | headless | browser-qa-jev --project-cwd . --url … --slug … --case "…" | |
codex | Codex gpt-6-astra / low + browser plugins | browser-qa-codex --project-cwd . --url … --slug … --case "…" | |||
claude | this agent via chrome-devtools MCP | steps below |
browser-qa-jev writes cases.md, REPORT.md, shots/, jev-run.json. browser-qa-codex writes codex-run.json / codex-result.json. Replay scripts are not produced in jev/codex mode; receipt + shots are the proof.
Preflight
- Read
.agents/qa/context.mdif present (routes, how to start the app, test-user role). No secrets. - Live browser: host-level
chrome-devtoolsMCP (mcp__chrome-devtools__*:list_pages,new_page,navigate_page,take_snapshot,take_screenshot), else MetaMCPchrome-devtools, else existing project Playwright. If the MCP reports no browser, runchrome-qa start(dedicated profile, fixed port, no consent dialog) and retry once. No new dependency. None → everybrowser-uicaseblocked. Do not writepass. WebFetch is not a browser.
Attached to the user's real Chrome (macOS --autoConnect, or --browserUrl): the server exposes every open tab. Create your own page with new_page, act only in pages you created, never snapshot, navigate or close other tabs, close yours when done, and prefer a dedicated QA profile for sensitive sites.
productionorunknown, or any side-effect case: do not run until the PM wroteauthorizedin the spawn prompt. Deny →blocked.- Auth: reuse a safe local fixture from context/docs. Never store prod/personal passwords, cookies, OTP. Disposable local users only if the repo already has a seed command.
Run
cdproject root. Slug from PM orqa-<YYYYMMDD-HHMM>.- Write or refresh
cases.mdfrom SPEC/acceptance. Do not invent URLs. qa-digest cases+qa-digest target. If REPORT exists, apply stale rules before changing statuses.- Create REPORT from
references/REPORT.mdwhen missing. All new cases start pending. - Each
browser-uicase: matching replay first; else write script, run once, shotsshots/TC-00N-<w>.png. - Supporting L0/L1 commands may be listed; they do not count as extra TC results.
- Last line to PM:
DONE .agents/qa/<slug>/REPORT.mdplus fail list and 3–5 shots.
NEVER (each with its alternative)
- Product Vue/TS/CSS and
docs/DESIGN.mdare not yours: report the defect in REPORT and leave the file. - A
browser-uicase ispassonly after a live browser run of that case; without a live browser it isblocked. - Work only in cases from
cases.md; for anything else add a## TC-00Nfirst. This is the Playwright-replay harness, so no Human mode,.ai-factoryor/aif-qa. - Credentials stay out of files, and the preview host stays out of scripts (
BASE_URLplaceholder). - Cleanup is read-only for shared state: no
rm,git resetor shared DB wipe.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingplugins/lane-stack/skills/browser-qa/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/vkirill/claude-lane-stack/browser-qa)