UI Test — Agentic UI Testing Skill

Test UI changes in a real browser. Your job is to try to break things, not confirm they work.

Three workflows:

• Diff-driven — analyze a git diff, test only what changed
• Exploratory — navigate the app, find bugs the developer didn't think about
• Parallel — fan out independent test groups across multiple Browserbase browsers

How Testing Works

The main agent coordinates — it plans test strategy, delegates to sub-agents, and merges results. Sub-agents do the actual browser testing.

Planning: multiple angles, then execute once

You MUST complete all three planning rounds yourself and output them before launching any sub-agents. Planning happens in your own response — it is NOT delegated to sub-agents. Do not skip ahead to execution.

Round 1 — Functional: What are the core user flows? What should work? Write out each test as: action → expected result.

Round 2 — Adversarial: Re-read Round 1. What did you miss? Think about: different user types/roles, error paths, empty states, race conditions, edge inputs (empty, huge, special chars, rapid clicks).

Round 3 — Coverage gaps: Re-read Rounds 1–2. What about: accessibility (axe-core, keyboard-only), mobile viewports, console errors, visual consistency with the rest of the app?

Deduplicate: Merge all three rounds into one numbered list of tests. Remove overlaps. Assign each test to a group (e.g. Group A, Group B).

Then execute once — launch one sub-agent per group. Each sub-agent receives its specific list of tests to run, nothing more. Sub-agents do not explore or plan — they execute assigned tests and report results.

Output the three rounds, the merged plan, and the group assignments in your response before calling any Agent tool.

Principles for splitting work

• Sub-agents run assigned tests, not open exploration. The main agent hands each sub-agent a specific numbered list of tests. Sub-agents do not plan, explore, or decide what to test — they execute the list and stop.
• The bottleneck is the slowest agent — split work so no single agent has a disproportionate share. Many small agents > few large ones.
• Size the effort to the change — a single component fix doesn't need many agents or many steps. A full-page redesign does. Let the scope of the diff drive the plan.
• No early stopping on failures — find as many bugs as possible within the assigned tests.

Giving sub-agents a step budget

The main agent MUST include an explicit browse step limit in every sub-agent prompt. Sub-agents do not self-limit — they will run until done unless told otherwise.

As a rough heuristic: ~25 steps for a few targeted checks, ~40 for a full page with functional + adversarial + a11y, ~75 for multiple pages or a broad category. Adjust based on what the assigned tests actually require — these are starting points, not rules.

As a rough heuristic: ~25 steps for a few targeted checks, ~40 for a full page with functional + adversarial + a11y, ~75 for multiple pages or a broad category. Adjust based on what the assigned tests actually require — these are starting points, not rules.

Every sub-agent prompt must include:

You have a budget of N browse steps (each `browse` command = 1 step). Count your steps as you go. When you reach N, stop immediately and report:
- STEP_PASS/STEP_FAIL for every test you completed
- STEP_SKIP||budget reached for every test you didn't get to

Do not retry or continue after hitting the budget.
Run only these tests: [numbered list from the merged plan]
Do not explore beyond the assigned tests.
Do NOT generate an HTML report or write any files. Return only step markers and your findings as text.

The main agent should NOT run browse commands itself (except to verify the dev server is up). All testing happens in sub-agents.

When a sub-agent hits its budget, the main agent accepts the partial results as-is. Do not re-run or retry the sub-agent. Include SKIPPED tests in the final report so the developer knows what wasn't covered.

Reporting

Every sub-agent reports back with:

Tests: 8 | Passed: 5 | Failed: 2 | Skipped: 1 | Pages visited: 2

The main agent merges into a final report with:

Tests: 20 | Passed: 14 | Failed: 4 | Skipped: 2 | Agents: 3 | Pass rate: 70%

Do not report "steps used" — browse command counts are implementation plumbing, not a meaningful metric for reviewers.

Testing Philosophy

You are an adversarial tester. Your goal is to find bugs, not prove correctness.

• Try to break every feature you test. Don't just check "does the button exist?" — click it twice rapidly, submit empty forms, paste 500 characters, press Escape mid-flow.
• Test what the developer didn't think about. Empty states, error recovery, keyboard-only navigation, mobile overflow.
• Every assertion must be evidence-based. Compare before/after snapshots. Check specific elements by ref. Never report PASS without concrete evidence from the accessibility tree or a deterministic check.
• Report failures with enough detail to reproduce. Include the exact action, what you expected, what you got, and a suggested fix.

Assertion Protocol

Every test step MUST produce a structured assertion. Do not write freeform "this looks good."

Step markers

For each test step, emit exactly one marker:

STEP_PASS||

or

STEP_FAIL|| → |

• step-id: short identifier like homepage-cta, form-validation-error, modal-cancel
• evidence: what you observed that proves the step passed (element ref, text content, URL, eval result)
• expected → actual: what you expected vs what you got
• screenshot-path: path to the saved screenshot (failures only — see Screenshot Capture below)

Screenshot Capture for Failures

Every STEP_FAIL MUST have an accompanying screenshot so the developer can see what went wrong visually.

When a test step fails:

# 1. Take a screenshot immediately after observing the failure
browse screenshot --path .context/ui-test-screenshots/.png

# If --path is not supported, take the screenshot and save manually:
browse screenshot
# The browse CLI will output the screenshot path — move/copy it:
cp /tmp/browse-screenshot-*.png .context/ui-test-screenshots/.png

Setup the screenshot directory at the start of any test run:

mkdir -p .context/ui-test-screenshots

Rules:

• File name = step-id (e.g., double-submit.png, axe-audit.png, modal-focus-trap.png)
• Store in .context/ui-test-screenshots/ — this directory is gitignored and accessible to the developer and other agents
• For parallel runs, include the session name: -.png (e.g., signup-double-submit.png)
• Take the screenshot at the moment of failure — capture the broken state, not after recovery
• For visual/layout bugs, also screenshot the baseline (working state) for comparison: -baseline.png

How to verify (in order of rigor)

1. Deterministic check (strongest) — browse eval returns structured data you can inspect. Examples: axe-core violation count, document.title, form field value, console error array, element count.
2. Snapshot element match — a specific element with a specific role and text exists in the accessibility tree. Check by ref: @0-12 button "Save". An element either exists in the tree or it doesn't.
3. Before/after comparison — snapshot before action, act, snapshot after. Verify the tree changed in the expected way (element appeared, disappeared, text changed).
4. Screenshot + visual judgment (weakest) — only for visual-only properties (color, spacing, layout) that the accessibility tree cannot capture. Always accompany with what specifically you're evaluating.

Before/after comparison pattern

This is the core verification loop. Use it for every interaction:

# 1. BEFORE: capture state
browse snapshot
# Record: what elements exist, their text, their refs

# 2. ACT: perform the interaction
browse click @0-12

# 3. AFTER: capture new state
browse snapshot
# Compare: what changed? What appeared? What disappeared?

# 4. ASSERT: emit marker based on comparison
# If dialog appeared: STEP_PASS|modal-open|dialog "Confirm" appeared at @0-20
# If nothing changed:
browse screenshot --path .context/ui-test-screenshots/modal-open.png
# STEP_FAIL|modal-open|expected dialog to appear → snapshot unchanged|.context/ui-test-screenshots/modal-open.png

Setup

which browse || npm install -g browse

Avoid permission fatigue

This skill runs many browse commands (snapshots, clicks, evals). To avoid approving each one, add browse to your allowed commands:

Add both patterns to .claude/settings.json (project-level) or ~/.claude/settings.json (user-level):

{
  "permissions": {
    "allow": [
      "Bash(browse:*)",
      "Bash(BROWSE_SESSION=*)"
    ]
  }
}

The first pattern covers plain browse commands. The second covers parallel sessions (BROWSE_SESSION=signup browse open ...). Both are needed to avoid approval prompts.

Mode Selection

Rule: If the target URL contains localhost or 127.0.0.1, pass --local on the first browse open.

Local Mode (default for localhost)

browse open http://localhost:3000 --local

browse open ... --local uses a clean isolated local browser by default, which is best for reproducible localhost QA runs.

Use local-mode variants only when needed:

• browse open --auto-connect — auto-discover an existing debuggable local Chrome. Use this only when the test explicitly needs existing local login/cookies/state.
• browse open --cdp — attach to a specific CDP target (explicit local browser attach).

Remote Mode (deployed sites via cookie-sync)

# Step 1: Sync cookies from local Chrome to Browserbase
node .claude/skills/cookie-sync/scripts/cookie-sync.mjs --domains your-app.com
# Output: Context ID: ctx_abc123

# Step 2: Open in remote mode with the synced context
SESSION_JSON="$(browse cloud sessions create --context-id ctx_abc123 --persist --keep-alive)"
SESSION_ID="$(echo "$SESSION_JSON" | jq -r .id)"
CONNECT_URL="$(echo "$SESSION_JSON" | jq -r .connectUrl)"

browse open https://staging.your-app.com --cdp "$CONNECT_URL"
browse snapshot
# ... run tests ...
browse stop
browse cloud sessions update "$SESSION_ID" --status REQUEST_RELEASE

Cookie-sync flags: --domains, --context, --verified, --proxy "City,ST,US"

Workflow A: Diff-Driven Testing

Phase 1: Analyze the diff

git diff --name-only HEAD~1          # or: git diff --name-only / git diff --name-only main...HEAD
git diff HEAD~1 --             # read actual changes

Categorize changed files:

Phase 2: Map files to URLs

Detect framework: cat package.json | grep -E '"(next|react|vue|nuxt|svelte|@sveltejs|angular|vite)"'

Phase 3: Ensure the right code is running

Before testing, verify the dev server is serving the code from the diff — not a stale branch.

If testing a PR or specific branch:

# Check what branch is currently checked out
git branch --show-current

# If it's not the PR branch, switch to it
git fetch origin  && git checkout 

# Install deps — the lockfile may differ between branches
yarn install  # or npm install / pnpm install

If the dev server was already running on a different branch, restart it after checkout.

Find a running dev server:

for port in 3000 3001 5173 4200 8080 8000 5000; do
  s=$(curl -s -o /dev/null -w "%{http_code}" "http://localhost:$port" 2>/dev/null)
  if [ "$s" != "000" ]; then echo "Dev server on port $port (HTTP $s)"; fi
done

If nothing found: tell the user to start their dev server.

Verify it actually renders: After browse open + browse snapshot, check that the accessibility tree contains real page content (navigation, headings, interactive elements) — not just an error overlay or empty body. Next.js dev servers can return HTTP 200 while showing a full-screen build error dialog. If the snapshot is empty or dominated by an error dialog, the server is broken — fix the build before testing.

Phase 4: Generate test plan

For each changed area, plan both happy path AND adversarial tests:

Test Plan (based on git diff)
=============================
Changed: src/components/SignupForm.tsx (added email validation)

1. [happy] Valid email submits successfully
   URL: http://localhost:3000/signup
   Steps: fill valid email → submit → verify success message appears

2. [adversarial] Invalid email shows error
   Steps: fill "not-an-email" → submit → verify error message appears

3. [adversarial] Empty form submission
   Steps: click submit without filling anything → verify error, no crash

4. [adversarial] XSS in email field
   Steps: fill "" → submit → verify sanitized/rejected

5. [adversarial] Rapid double-submit
   Steps: click submit twice quickly → verify no duplicate submission

6. [adversarial] Keyboard-only flow
   Steps: Tab to email → type → Tab to submit → Enter → verify success

Phase 5: Execute tests

browse stop 2>/dev/null
mkdir -p .context/ui-test-screenshots
# localhost/default QA → clean, reproducible local run
browse open http://localhost:3000 --local

For each test, follow the before/after pattern:

# Navigate
browse open http://localhost:3000/path --local
browse wait load

# BEFORE snapshot
browse snapshot
# Note the current state: elements, refs, text

# ACT
browse click @0-ref
# or: browse fill "selector" "value"
# or: browse type "text"
# or: browse press Enter

# AFTER snapshot
browse snapshot
# Compare against BEFORE: what changed?

# ASSERT with marker
# STEP_PASS|step-id|evidence  OR  STEP_FAIL|step-id|expected → actual

Phase 6: Report results

## UI Test Results

### STEP_PASS|valid-email-submit|status "Thanks!" appeared at @0-42 after submit
- URL: http://localhost:3000/signup
- Before: form with email input @0-3, submit button @0-7
- Action: filled "[email protected]", clicked @0-7
- After: form replaced by status element with "Thanks! We'll be in touch."

### STEP_FAIL|double-submit|expected single submission → form submitted twice|.context/ui-test-screenshots/double-submit.png
- URL: http://localhost:3000/signup
- Before: form with submit button @0-7
- Action: clicked @0-7 twice rapidly
- After: two success toasts appeared, suggesting duplicate submission
- Screenshot: .context/ui-test-screenshots/double-submit.png
- Suggestion: disable submit button after first click, or debounce the handler

---
**Summary: 4/6 passed, 2 failed**
Failed: double-submit, xss-sanitization

Screenshots saved to `.context/ui-test-screenshots/` — open any failed step's screenshot to see the broken state.

Always browse stop when done.

Phase 7: Generate HTML report

After producing the text report, generate a standalone HTML report that a reviewer can open in a browser. The report embeds screenshots inline (base64) so it works as a single file — no external dependencies.

Why: Text reports are good for the agent conversation, but reviewers (PMs, designers, other engineers) want a visual artifact they can open, scan, and share. Screenshots inline make failures immediately obvious.

How to generate

1. Read the HTML template at references/report-template.html
2. Build the report by replacing the template placeholders with actual test data: