{"id":41008,"plugin_id":"plugin_asdk_app_6a53f7d79b5c81919a02e9814eb9f070","kind":"skill","collection_source":"plugin_package","comparison_source":null,"observed_at":"2026-10-24T12:56:56.429Z","digest":"7728b92902ae77b34d9d8952806d8f606e483f755a1a5d9543ee454e9ddba3ad","against":null,"payload":{"description":"Builds a Supademo demo from uploaded video or prepared screenshots, then adds copy and structure. Use for a recording-based demo; explain when a requested screenshot sequence needs separate capture or extraction.","included_files":[{"relative_path":"agents/openai.yaml","size_in_bytes":217}],"name":"video-to-demo-converter","skill_md_contents":"---\nname: video-to-demo-converter\ndescription: Builds a Supademo demo from uploaded video or prepared screenshots, then adds copy and structure. Use for a recording-based demo; explain when a requested screenshot sequence needs separate capture or extraction.\n---\n\n# Video to Demo Converter\n\nTurn a recording or prepared media sequence into a demo using the media handles Supademo actually supports.\n\n## Workflow\n\n1. **Choose the output.** One uploaded video becomes one video step. A screenshot walkthrough needs separate image assets, one per step. The MCP tools do not automatically split a recording into screenshot steps. If the user needs that output, prepare the capture/extraction plan and use a separately available, authorized capture workflow, or ask for the screenshots.\n2. **Get trusted media handles.** Use conversation handles if present. For local files, call `create_upload_job`, provide its upload URL, and call `get_upload_job` after upload until processing completes. Use returned `videoId` for a completed video, or image URL, width, and height for each image. For explicitly requested library reuse, call `list_media_library_items`: a video's `itemId` is its `videoId`; image items provide `url`, `width`, and `height`. If an image lacks positive dimensions, request a completed upload with those values rather than inventing them.\n3. **Build.** For a new demo, call `create_demo_from_uploaded_media` with the ordered media array. Map selected library videos to `{ type: \"video\", videoId: itemId }` and images to `{ type: \"image\", imageUrl: url, width, height }`. Each item becomes one regular step in that order. Preserve the returned demo and step IDs. To add library items to an existing demo instead, resolve the target with `get_demo`, then call `create_step_from_media_library_item` with that `demoId`, the selected `itemId`, and its `type`. That tool adds a step; it does not create a demo or return a media handle.\n4. **Edit the result.** Read `get_demo`. Add clear step copy with `update_steps` and appropriate hotspots with `create_hotspots`. Create new task-based chapter steps with `create_steps` using `type: \"chapter\"`, `chapterType: \"Link\"`, the title in `chapter.heading`, and the planned 1-based `position`; use `update_chapters` to edit existing chapter step IDs. Use `reorder_steps` or `delete_steps` only for changes included in the user's request. There is no generic step-merge tool. Removing a step does not trim dead air inside a video.\n5. **Handle audio explicitly.** Do not claim the source narration was removed or restored by conversion. Inspect the resulting video/audio. If new narration is requested, use `voiceover-producer`; adding generated voiceover is separate from editing or muting source video audio.\n6. **Verify.** Use `get_demo` and `demo-qa-checklist`. Check the actual media, step order, copy, links, and playback when browser access is available. Publish only when publication is part of the authorized task.\n\n## Source review\n\nInspect for private data, notifications, and unreadable captures before sharing. Explain any required recapture or redaction, and distinguish completed edits from work the available tools cannot perform.\n\n## Supademo tool conventions\n\nRead [the tool conventions](../supademo-onboarding/references/tool-use.md) before calling Supademo tools for this workflow.\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}