Skip to content

Interactive heading — evidence and limits ​

Preparation checkpoint, 5 October 2026. The original S-11 asks whether Interactive Demo, Product Configurator and Interactive Experience should share a heading. Their type-name approval and their physical registration are different matters. The heading is now approved; the checks below do not establish schema changes or working interactive analysis.

Recovered history ​

The August ratification reply approves individual content types while reserving their family headings. RUL-078 records that distinction. The preserved proposed-content-types page lists Interactive Demo and Product Configurator with product communication and Interactive Experience with event, retail and interactive work. Its historical banner warns that these families are superseded and its implementation statements are dated.

The original S-11 card asks only about the heading. Its database explanation says Interactive Demo and Product Configurator are in the product-communication family. The selected live records below contradict that explanation. The original card is preserved as historical source; the current preparation draft corrects the premise rather than overwriting its history.

At preparation, neither the kiosk placement nor this broader heading was answered. Both were subsequently approved on 5 October in separate replies. See the heading decision and kiosk decision. The dated checks below remain limited implementation evidence.

Selected live checks ​

The repository receipt docs/reference/2026-10/2026-10-04/interactive-heading-preflight.json retains exact queries and raw results. Schema was inspected before querying selected fields.

  • The selected registry query returned Interactive Experience and Product Configurator under event_retail_interactive. It returned no interactive_demo row. Both returned type rows have no inclusion or exclusion rules in those selected fields.
  • The selected content-type alias query returned no matching interactive or configurator alias. This is not an exhaustive audit of all normalization or alias paths.
  • The grouped query for undeleted pieces with the three exact type values, or with the exact interactive media value, returned an empty result. That establishes zero matching rows under those predicates. It does not establish that the library contains no interactive work, screenshots or recordings classified another way.

Registration reconciliation is engineering work. It must not reopen the already-approved names or manufacture type approval from a heading answer.

Checked example ​

Opened the Storylane demonstration, linked from its official recording documentation. Its title was “How to capture HTML demo.” The embedded player exposed a hotspot button. Clicking it changed the displayed captured-image URL from 17wib75r6vxfdbgctvxedtysyipn.png to 4m4pm00zfp6poro5aw4eedhkkn0y.png. The resulting screenshot showed step 2/8, a Next control and instructions about the capture button.

This directly establishes a user-controlled step change in this example. It does not prove the whole eight-step demonstration, underlying product functionality or arbitrary interaction replay. The screenshot-based example is intentionally not described as a functioning copy of every product control. No account was created, software installed or capture started.

Local consumer trace ​

The source-insert helper accepts a media-format argument and writes it to the content record; accepting a value does not implement its analysis. The admin prompt-routing table supplies a literal MOCK_STAGE_ROUTING row for interactive work, and the prompt landing page passes that mock to its overview. That display is not runtime evidence.

The local orchestrator's selected analysis dispatch recognizes video, text and image. Other values take the compatibility branch to video when a media path exists, otherwise text. It then calls the selected text, image or video handler. This is a specific local-code limitation to investigate before claiming interactive analysis. No deployed-file equality, real execution or complete ingest-to-interface trace was verified for this subject.

The older batch analyzer excerpt is frame/video-oriented. Its presence alone does not establish which production entry point handles a given capture. The full audit must establish routing reachability, input gates, composed prompts, output validation, writes, publication and readers before specifying a repair.

Bounded challenge and remaining work ​

Retaining the proposed heading improves discovery across product and non-product experiences. Folding product examples into Product & Feature simplifies prominent navigation but leaves other interactive work dependent on context. A shared heading need not dictate a physical hierarchy, force exclusive membership or treat every clickable page as an experience.

The full audit still owes exact type boundaries, capture/representation identity, evidence of observed states, missing or withdrawn experiences, permissions, versions, human corrections, query parity and monitoring behavior. Scheduling and cost controls remain part of that audit. This preparation changes no public API, production data, prompt or automation setting.

Last reviewed 5 October 2026.

The BrandTrackers domain model. Source: git markdown, drift-checked against the live DB.