The observed fact is that users can choose an app from a gallery or describe a task, upload photos and clips, and continue from earlier results. A second observed fact is that the listed apps span product ads, headshots, short films, photo animation, relighting, camera changes, outfit swaps, lip-sync, and soundtracks; the page also names a Pika MCP server and multiple skills.
Mechanism: Packaging task-specific creative apps and skills within a chat interface provides discoverable routes into specialized operations while retaining conversational context. Rather than exposing a single undifferentiated generation prompt, the listing presents a gallery of applications and task-related skills that can route a user toward a specific creative operation while retaining conversational context. The source supports a workflow claim, not a claim about model quality, adoption, or broad autonomy. Evidence boundary: The public listing does not establish launch timing, availability, whether every listed capability works as described in practice, model quality, adoption, or user outcomes.
Why now: the listing makes the chat surface itself a distribution and workflow layer for a specialized creative suite. Implication: Application providers can make capabilities discoverable as bounded tasks inside an assistant rather than requiring users to start in a separate studio. Watch condition: whether the integration supports reliable handoff between these task apps and preserves user control over inputs and generated assets. This starts a distinct analytical record because no prior New Runtime record is supplied for this specific Pika-in-ChatGPT workflow.