Functional Motion Review
Assess whether motion helps a named reader task. Review only; this skill does not authorize code changes, installs, analytics, user recruitment, publication, or deployment.
Read motion terms and evidence checks before assigning motion roles or intensity, judging replay time, or making an accessibility finding.
Review
- Identify the reader's question and the intended benefit. Separate a proposed benefit from observed comprehension findings.
- Trace each displayed object to its authority, version, timestamp, and evidence kind. Domain state must commit independently of animation completion. Keep live observations, examples, simulations, predictions, and retrospectives visibly distinct.
- Name the relationship the motion expresses. A sequence or connector cannot prove causation. Use the reference to distinguish chronology, correlation, modeled causality, reported causality, revision, and provenance.
- Assign a local role and intensity using the reference. These terms describe a review; they are not adoption targets. Ordinary reading can remain static.
- Inspect complete initial content, document order, native links and disclosures, keyboard operation, focus, non-color labels, reduced motion, and interruption. The catalog prefers a static start when motion preference is unknown and no automatic replay when a hidden page returns. Keep those preferences distinct from WCAG requirements.
- Compare exact baseline and candidate builds, dependencies, bundle sizes, layout shifts, long tasks, and repeatable lab runs under the same device settings. A Lighthouse navigation result does not measure interaction latency or field Core Web Vitals.
- Examine consented, counterbalanced task findings when available: errors, completion times, and confusion. Preference and dwell time do not establish comprehension. Keep absent participant evidence explicitly absent; do not propose production instrumentation to fill the gap.
Route detailed reviews
- Use
accessibility-review for criterion-level findings and assistive-technology checks.
- Use
performance-scalability-diagnosis for measured performance regressions.
- Use
api-contract-compatibility-review when a presentation change touches producer or consumer contracts.
- Use
authority-boundary-review if the evidence authority itself is unresolved.
If a companion is unavailable, report that limit and use its primary authority without inventing a completed review. Do not duplicate its full procedure here.
Stop and report
Reject fake live activity, hidden essential content, unlabeled synthetic outcomes, playback that cannot be interrupted, and analytics scope expansion. Do not repair a domain contract to accommodate animation. Keep each project's terms, runtime, and visual identity local; reuse a recipe only after multiple pilots demonstrate the same behavior. Missing evidence is not an engagement improvement or an accessibility conformance claim.
Output
Lead with the supported finding about the reader task. Put a blocking failure or incomplete evidence before any proposed benefit. Preserve the requested detail, failed checks, uncertainty, and relevant unfinished work.
Then report the intended benefit, evidence and domain authority, motion role and intensity, static and keyboard behavior, reduced motion and interruption, bundle and performance evidence, comprehension findings, risks, and missing evidence. Label each conclusion as repository inspection, browser observation, automated test, participant finding, or recommendation. Include the date and exact reviewed revision. Give a next action only when one is needed.
WCAG 2.3.3 interaction animation is Level AAA. Pause, Stop, Hide (2.2.2) is Level A with separate conditions. Use the reference to assess those conditions; do not describe every motion preference as a universal AA requirement.
1---2name: functional-motion-review3description: Review whether interface motion clarifies evidence, sequence, comparison, or reader actions. Use for animated editorial and data experiences; not for routine styling, implementing animation, or claiming engagement from visual polish.4license: MIT5---67# Functional Motion Review89Assess whether motion helps a named reader task. Review only; this skill does not authorize code changes, installs, analytics, user recruitment, publication, or deployment.1011Read [motion terms and evidence checks](references/motion-evidence.md) before assigning motion roles or intensity, judging replay time, or making an accessibility finding.1213## Review14151. Identify the reader's question and the intended benefit. Separate a proposed benefit from observed comprehension findings.162. Trace each displayed object to its authority, version, timestamp, and evidence kind. Domain state must commit independently of animation completion. Keep live observations, examples, simulations, predictions, and retrospectives visibly distinct.173. Name the relationship the motion expresses. A sequence or connector cannot prove causation. Use the reference to distinguish chronology, correlation, modeled causality, reported causality, revision, and provenance.184. Assign a local role and intensity using the reference. These terms describe a review; they are not adoption targets. Ordinary reading can remain static.195. Inspect complete initial content, document order, native links and disclosures, keyboard operation, focus, non-color labels, reduced motion, and interruption. The catalog prefers a static start when motion preference is unknown and no automatic replay when a hidden page returns. Keep those preferences distinct from WCAG requirements.206. Compare exact baseline and candidate builds, dependencies, bundle sizes, layout shifts, long tasks, and repeatable lab runs under the same device settings. A Lighthouse navigation result does not measure interaction latency or field Core Web Vitals.217. Examine consented, counterbalanced task findings when available: errors, completion times, and confusion. Preference and dwell time do not establish comprehension. Keep absent participant evidence explicitly absent; do not propose production instrumentation to fill the gap.2223## Route detailed reviews2425- Use `accessibility-review` for criterion-level findings and assistive-technology checks.26- Use `performance-scalability-diagnosis` for measured performance regressions.27- Use `api-contract-compatibility-review` when a presentation change touches producer or consumer contracts.28- Use `authority-boundary-review` if the evidence authority itself is unresolved.2930If a companion is unavailable, report that limit and use its primary authority without inventing a completed review. Do not duplicate its full procedure here.3132## Stop and report3334Reject fake live activity, hidden essential content, unlabeled synthetic outcomes, playback that cannot be interrupted, and analytics scope expansion. Do not repair a domain contract to accommodate animation. Keep each project's terms, runtime, and visual identity local; reuse a recipe only after multiple pilots demonstrate the same behavior. Missing evidence is not an engagement improvement or an accessibility conformance claim.3536## Output3738Lead with the supported finding about the reader task. Put a blocking failure or incomplete evidence before any proposed benefit. Preserve the requested detail, failed checks, uncertainty, and relevant unfinished work.3940Then report the intended benefit, evidence and domain authority, motion role and intensity, static and keyboard behavior, reduced motion and interruption, bundle and performance evidence, comprehension findings, risks, and missing evidence. Label each conclusion as repository inspection, browser observation, automated test, participant finding, or recommendation. Include the date and exact reviewed revision. Give a next action only when one is needed.4142WCAG 2.3.3 interaction animation is Level AAA. Pause, Stop, Hide (2.2.2) is Level A with separate conditions. Use the reference to assess those conditions; do not describe every motion preference as a universal AA requirement.
JovaniPink