← Browse

@jinning6/browser-actionability-debug

A

Diagnose browser UI tests or real browser checks where an element appears visible but cannot be clicked, copied, hovered, or focused. Use for Playwright/Chrome/browser automation failures involving overlays, splash screens, opacity transitions, pointer-events, elementFromPoint mismatches, clipboard assertions, or differences between "visible" and actually actionable UI.

skillclaude

Install

agr install @jinning6/browser-actionability-debug --target claude

Writes 1 file into .claude/skills/, pinned to git-80d7dad1.

  • .claude/skills/browser-actionability-debug/SKILL.md

Document


name: browser-actionability-debug description: Diagnose browser UI tests or real browser checks where an element appears visible but cannot be clicked, copied, hovered, or focused. Use for Playwright/Chrome/browser automation failures involving overlays, splash screens, opacity transitions, pointer-events, elementFromPoint mismatches, clipboard assertions, or differences between "visible" and actually actionable UI.

Browser Actionability Debug

Workflow

  1. Reproduce the failure with a real browser action, not only DOM existence.

    • Prefer normal locator.click() first so the browser reports what intercepts the pointer.
    • Read the full call log. If it names an intercepting element, treat that as evidence.
  2. Separate visibility from actionability.

    • visible can still be blocked by an overlay, opacity: 0, parent pointer-events: none, a splash screen, or an element with higher stacking order.
    • Inspect elementFromPoint() at the target center and compare it to the intended button/input.
    • Capture computed opacity, visibility, display, pointerEvents, and bounding boxes for the target and its stable ancestors.
  3. Check lifecycle gates before changing CSS.

    • If a splash/loading/transition component exists, wait for the lifecycle signal that makes the app interactive, such as overlay detached or app shell pointer-events: auto.
    • Do not fix a test by using force: true unless the user specifically needs to bypass hit testing. It can hide real UX bugs.
  4. Fix at the correct layer.

    • If the overlay should still be active, update the test to wait for the overlay to detach.
    • If a decorative canvas or visual layer should never capture input, set pointer-events: none on that visual layer.
    • If app content is intentionally disabled until boot, do not make underlying controls clickable before boot unless that is a product decision.
  5. Verify clipboard behavior cross-platform.

    • On Windows, browser clipboard reads may normalize line endings to CRLF. Normalize \r\n to \n for semantic multi-line copy assertions.
    • Still assert exact command text after line-ending normalization.

Diagnostic Snippet

Use this inside page.evaluate() when a click target exists but is not actionable:

const target = document.querySelector('[aria-label="Copy Noosphere share post"]');
const rect = target?.getBoundingClientRect();
const top = rect
  ? document.elementFromPoint(rect.left + rect.width / 2, rect.top + rect.height / 2)
  : null;

return {
  target: target ? getComputedStyle(target).cssText : null,
  topTag: top?.tagName,
  topClass: String(top?.className || ''),
  topAria: top?.getAttribute?.('aria-label') || null,
  targetRect: rect ? {
    x: rect.x,
    y: rect.y,
    width: rect.width,
    height: rect.height,
  } : null,
};

Completion Standard

  • The original browser action succeeds without force: true.
  • The target is within the tested viewport.
  • elementFromPoint() at the target center resolves to the target or a child of it.
  • Clipboard assertions normalize OS line endings but otherwise match exactly.
  • Dev servers and temporary browser dependencies are stopped or removed after verification.

Trustgrade A

  • passBody integrity

    Whether the stored document is plausibly the kind of file the artifact declares, rather than something fetched by mistake.

  • passType matchnot applicable to this artifact type

    Whether the artifact is really the kind of thing its metadata claims it is.

  • passFreshness

    How long since the source repository was last pushed to.

  • passPrompt injection

    Scans the artifact's own text for instructions aimed at your agent rather than at you.

  • passLicense

    Whether the source repository declares an SPDX license permissive enough to redistribute.

How the grade is calculated

Each check contributes 0 points when it passes, 1 when it warns, and 2 when it fails. The total maps to a letter:

  • Aevery check passed
  • Bone warning
  • Ctwo warnings
  • Dprompt injection or body integrity failed, or three warnings
  • Fone of those failed, and something else is wrong

These are automated hygiene checks, not a security audit, and not a dependency or vulnerability scan. A grade of A means nothing was flagged — not that the artifact is safe.

Versions

  • git-80d7dad158de2026-07-31