{
  "authorship": "Five self-authored fictional cases, not vendor results or actual observations.",
  "cases": [
    {
      "id": 1,
      "text": " 1. A loop behind a stationary verse   Fictional setup:  Mara places a verse on a landing page over a decorative cloud loop. In the imagined design, the loop starts when the page loads and repeats while the verse stays visible. The only obvious pause button belongs to a separate song player. Her draft inventory has two rows, cloud backdrop and song player, because one button label cannot establish control over both.   Proposed decision:  Give the cloud row its own stable-state action or connect it to a clearly described shared motion control. The review entry should say which layer stopped and whether the entire verse stayed readable. A muted speaker icon would answer a different question. If Mara cannot determine the duration or trigger in the delivered page, those cells remain pending rather than receiving a pass from the mockup.   Handoff:  Attach the exact backdrop revision and the intended stationary image. Ask the reviewer to activate the control, move on to a lyric link and note whether the backdrop starts again. This is an authored test plan, not an observed successful fix. "
    },
    {
      "id": 2,
      "text": " 2. An animated sidebar outside the video   Fictional setup:  Owen prepares a lyric video with a normal playback pause control. Beside it, a promotional strip scrolls automatically through project names. Pausing the video in his imagined layout leaves the strip moving. His initial note, \u201cplayer has pause,\u201d covers only one row of the inventory and cannot describe the surrounding page.   Proposed decision:  Treat the strip as another layer with its own trigger, duration and control. Preserve a readable way to reach the project links if the strip is hidden. If a stationary list replaces it, compare the destinations with the original approved list rather than assuming that freezing one item preserves all choices.   Handoff:  Owen records two requested outcomes: stop the relevant strip motion and retain access to the project destinations. He leaves both outcomes unverified until someone checks the published template. A change to the video export alone does not close the sidebar task. The example illustrates review scope, not a claim about any real publisher's controls. "
    },
    {
      "id": 3,
      "text": " 3. Motion pauses only while focus stays on a button   Fictional setup:  In Priya's prototype, moving artwork becomes still while a certain control has focus. As soon as she would navigate onward to read a related lyric note, the artwork would resume. That behavior is stipulated for this fictional exercise; it was not observed in a browser test.   Proposed decision:  Specify a persistent reading state that does not require the reader to keep focus on the artwork control. Keep a separate record of activation, moving onward and explicit resumption. The reviewer needs to know whether leaving the control undoes the action, because reaching another link is part of the reading task rather than an accidental interruption.   Handoff:  Priya writes an expected sequence: activate the stable state, visit the next relevant control, read the verse, then intentionally resume if desired. Actual keys, focus destination and observed motion remain blank in the downloadable record. She does not mark the prototype accessible simply because one frozen frame exists. A proposed interaction and an actual completed walkthrough belong in different columns. "
    },
    {
      "id": 4,
      "text": " 4. Several moving layers with different controls   Fictional setup:  Ellis combines an animated backdrop, a rotating album card and a video player. The imagined \u201cPause video\u201d action stops only the player. A separate switch freezes the backdrop but leaves the card rotating. The design might be understandable with separate controls, but a broad label such as \u201cStop all motion\u201d would promise more than the described behavior supplies.   Proposed decision:  Create one inventory row per layer and map each action to its affected rows. If a shared control is proposed, specify its scope and test each row after activation. Preserve the individual playback controls when they serve a different purpose. Do not count three buttons as evidence that all three moving layers are covered.   Handoff:  Ellis supplies a simple mapping: control name, affected layer, intended stable state and pending observation. When the rotating card is replaced later, only the affected mapping is revised, but the shared control is checked again. This invented example adds a reviewable ownership record without claiming a universal player shortcut or measured product behavior. "
    },
    {
      "id": 5,
      "text": " 5. A stationary rendition loses meaningful information   Fictional setup:  Noor's visual sequence includes an on-screen dedication that appears between verses. Her proposed static fallback uses the opening artwork, which does not contain that dedication. The invented fallback is calmer, but its content inventory is incomplete. The task is therefore not finished merely by choosing a still image.   Proposed decision:  List the meaningful information that the stable rendition must preserve. Noor can plan a stationary note or an accompanying descriptive transcript for the dedication, then review wording and association with the song. Decorative movement and meaningful visual information receive separate decisions. She should not label all movement essential simply because the existing export contains it.   Handoff:  Save the approved dedication text, its relationship to the song and the proposed alternative location. Ask a reviewer to confirm that a reader can find it without restarting the distracting layer. Leave discovery and wording checks pending until observed. This scenario demonstrates content preservation in an authored workflow; it does not establish that an alternative is sufficient for every accessibility requirement. "
    }
  ]
}