Appropriate use
Use for video, image, audio, webpages, embeds, playlists, channels, multi-zone layouts, and digital-signage playback.
Scope and intent
A controlled workflow specification
This page defines the behaviour, controls, evidence, and outputs required from the skill. It does not claim that a language model can enforce security or correctness by wording alone.
Non-goals
Do not use to infer media correctness from a single screenshot or to test licensed customer content outside approved handling.
- 8
- versioned workflow phases
- 4
- explicit approval gates
- 6
- defined result states
- 10
- versioned output artifacts
Material changes to the source, environment, identity, approved actions, or expected behaviour invalidate dependent approvals and require revalidation.
Workflow
The 8 phases
Each phase has explicit inputs and outputs. Approval gates are hard stops bound to exact versions and hashes.
Phase 1: Playback contract preflight Gate 1 · approve expectation
Resolve content type, source, codec, resolution, frame rate, audio, duration, orientation, layout, schedule, looping, transition, and supported platform expectations.
media metadata · layout · timing · platform · expected behaviour
Phase 2: Build, display, and device attestation
Verify player build, OS or firmware, browser engine, device model, display resolution, refresh rate, orientation, audio route, decoder capability, and network.
artifact · hardware · display · decoder · network
Phase 3: Media corpus and scenario design
Select synthetic and approved representative content for startup, loop, transition, zone, embed, long duration, interruption, update, and unsupported-format behaviour.
corpus IDs · boundary files · known-good files · negative files
Phase 4: Content and side-effect approval Gate 2 · approve actions
Approve uploads, downloads, streaming endpoints, third-party embeds, audio output, schedules, remote commands, and storage use.
content rights · endpoint allowlist · audio · storage · cleanup
Phase 5: Playback execution
Measure startup, rendering, frame continuity, audio, aspect ratio, crop, zone placement, transition timing, and loop boundaries with video, logs, and decoder metrics.
time to first frame · dropped frames · A/V · layout · loop
Phase 6: Interruption and recovery
Test network loss, source failure, restart, display disconnect, sleep, cache update, playlist change, and recovery without treating temporary output as long-term stability.
interruption point · fallback · retry · recovery time
Phase 7: Finding and stability assessment Gate 3 · confirm findings
Classify startup failure, black screen, freeze, stutter, decode error, layout error, audio issue, sync issue, embed restriction, environment issue, or unsupported content.
frequency · duration · platform scope · confidence
Phase 8: Publish matrix and restore Gate 4 · publish
Publish platform-content outcomes, evidence, known limitations, recommended transcode or configuration guidance, and cleanup uploaded test content.
matrix · findings · guidance · cleanup · regression proposal
Approval contract
Every gate approves an exact decision object
An approval cannot be reused after its environment, source, action plan, or approved document changes.
| Gate | Decision | Required binding |
|---|---|---|
| 1 | Gate 1 · approve expectation | Approval stores approver, role, timestamp, context hash, approved document hash, expiry, and invalidation state. |
| 2 | Gate 2 · approve actions | Approval stores approver, role, timestamp, context hash, approved document hash, expiry, and invalidation state. |
| 3 | Gate 3 · confirm findings | Approval stores approver, role, timestamp, context hash, approved document hash, expiry, and invalidation state. |
| 4 | Gate 4 · publish | Approval stores approver, role, timestamp, context hash, approved document hash, expiry, and invalidation state. |
Result model
Defined states prevent vague conclusions
Attempt outcomes remain separate from these workflow results. Errors, cancellations, and invalid context are never silently converted into a pass.
PASS
Playback meets the approved visual, audio, timing, layout, and stability expectation.
PLAYBACK FAIL
Content does not start, continue, loop, transition, or recover as required.
QUALITY DEGRADED
Playback continues but frame, audio, timing, or visual quality is below the approved threshold.
PLATFORM-SPECIFIC
The failure follows a verified platform, firmware, decoder, or browser-engine condition.
UNSUPPORTED CONTENT
The media violates the documented supported-content contract.
INCONCLUSIVE
Observation duration or evidence cannot establish the claimed playback behaviour.
Evidence contract
Evidence must directly support the claim
Evidence has provenance, capture context, privacy classification, retention, and integrity metadata. A convenient artifact is not automatically sufficient proof.
| Claim | Primary evidence | Required validation |
|---|---|---|
| Media identity | File hash and media metadata | Codec, profile, resolution, frame rate, audio, and duration are known. |
| Startup | Video plus timestamped player logs | Time to first frame and visible output are measured. |
| Continuity | Frame/decoder metrics and observation video | Freeze, stutter, drops, and recovery are detectable. |
| Layout | Full-display screenshot or video | Zones, crop, aspect ratio, orientation, and safe area are visible. |
| Audio | Route state and approved recording or meter | Presence, sync, level, and mute behaviour are verified. |
| Long run | Periodic health samples and logs | Actual duration, interruptions, and monitoring gaps are recorded. |
Trust and safety
Controls that must be enforced outside the prompt
These controls belong in the runtime, connector permissions, sandbox, renderer, storage layer, and review process.
Runtime control
Content provenance
Test files have known rights, hashes, and expected decoder behaviour.
Runtime control
No screenshot-only pass
Motion, timing, looping, audio, and stability require time-based evidence.
Runtime control
Platform identity
Model, firmware, browser engine, and hardware revision are recorded.
Runtime control
External embed policy
Third-party terms, authentication, and endpoint allowlists are respected.
Runtime control
Audio safety
Volume and physical environment are controlled.
Runtime control
Observation honesty
A five-minute run cannot claim all-day stability.
Decision policy
Ordered outcomes, not averaged pass counts
The first applicable high-severity outcome takes precedence. A manual override is recorded separately and never rewrites the calculated result.
INVALID TEST
Content, build, display, or evidence identity cannot be established.
REJECT
Required playback or recovery fails.
BLOCKED
Required platform, content, or observation duration is unavailable.
PASS WITH LIMITS
Required playback passes with accepted platform or quality limitations.
PASS
Required playback and stability criteria pass for the tested matrix.
Known limitations
Uncertainty and coverage limits remain visible
The workflow documents what it cannot prove and what further evidence would change confidence.
- Limit 1. Browser embeds can change independently of the player release.
- Limit 2. Hardware decoders may behave differently for files with similar visible properties.
- Limit 3. Display processing can add crop, scaling, latency, or colour changes.
- Limit 4. Network streaming tests depend on CDN and origin behaviour.
- Limit 5. Long-duration freezes may require telemetry rather than continuous video.
- Limit 6. Copyright or customer privacy may limit sharing source files.
Outputs
Versioned records for review and continuation
Human summaries are generated from validated structured records. They are not maintained as separate, drifting sources of truth.
playback-context.jsonVersioned output produced by the workflow.
media-corpus.yamlVersioned output produced by the workflow.
platform-display-map.jsonVersioned output produced by the workflow.
scenario-plan.mdVersioned output produced by the workflow.
attempt-ledger.jsonlVersioned output produced by the workflow.
decoder-metrics/Versioned output produced by the workflow.
evidence-manifest.jsonVersioned output produced by the workflow.
playback-matrix.csvVersioned output produced by the workflow.
cleanup-report.mdVersioned output produced by the workflow.
playback-report.htmlVersioned output produced by the workflow.
artifacts/qa/playback-qa/<run-id>/ ├── playback-context.json ├── media-corpus.yaml ├── platform-display-map.json ├── scenario-plan.md ├── attempt-ledger.jsonl ├── decoder-metrics/ ├── evidence-manifest.json ├── playback-matrix.csv ├── cleanup-report.md └── playback-report.html
Example
Example request and controlled output
Input
Test a 4K YouTube embed and a local MP4 on SCOS, PIXI, Tizen 6.5, and webOS 6. Check startup, 30-minute playback, loop, audio, and recovery after network loss.
Output excerpt
Playback decision: PASS WITH LIMITS Local MP4: Passed on all four platforms. YouTube: - SCOS: PASS - PIXI: PASS - Tizen 6.5: QUALITY DEGRADED after reconnect - webOS 6: PLAYBACK FAIL on second launch Limit: The 30-minute run does not establish all-day stability. Next: Route the webOS repeat-launch failure to bug triage and run a focused Tizen reconnect session.
Production readiness
Required before operational use
The written skill is only one layer. The repository, runtime, connectors, evidence storage, and publication path must implement these requirements.
- Required. Tool, filesystem, repository, command, secret, and network permissions are enforced outside the model.
- Required. Every source, environment, build, plan, and approval has a version or content hash.
- Required. Untrusted values are escaped, length-bounded, path-safe, and separated from executable instructions.
- Required. Attempt outcomes, scenario verdicts, workflow decisions, and publication status remain separate.
- Required. Raw evidence, publishable evidence, access control, retention, redaction, and deletion are defined.
- Required. Concurrent runs use unique IDs, locks, idempotency keys, and conflict-aware updates.
- Required. Material changes invalidate dependent approvals and outcomes.
- Required. The repository validator passes before the skill is considered production-ready.
In one line
/playback-qa
Validates media startup, decoding, layout, looping, transitions, audio, synchronization, interruption, recovery, long-running stability, and platform-specific rendering.