Update to anyVideo

{
  "description": "Guide a comparison of two YouTube explanations of the same topic using a shared viewing checklist and the user's reported takeaways. Use when the user asks to compare explainers, teaching approaches, examples, or understanding across videos. Keep metadata-based selection separate from content comparison. Do not use for automatic video summaries, unsupported claims about a video's viewpoint, or a simple single-video recommendation.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 275
    }
  ],
  "name": "compare-explainers",
  "skill_md_contents": "---\nname: compare-explainers\ndescription: Guide a comparison of two YouTube explanations of the same topic using a shared viewing checklist and the user's reported takeaways. Use when the user asks to compare explainers, teaching approaches, examples, or understanding across videos. Keep metadata-based selection separate from content comparison. Do not use for automatic video summaries, unsupported claims about a video's viewpoint, or a simple single-video recommendation.\n---\n\n# Compare Explainers\n\n## Inputs\n\nEstablish the topic, the user's starting level, and what they want the comparison\nto resolve. Reuse supplied links or an existing selection. Default to two videos;\ninclude a third only when requested or needed to address a specific remaining gap.\n\n## Workflow\n\n1. Define a common comparison question and a short checklist: the main idea,\n   examples used, assumptions or prerequisites, and what remains unclear.\n   Adapt these dimensions to the user's goal.\n2. Discover missing candidates with `search-videos({ query, reason? })` and read\n   `results`. Select videos covering the same question, preferably with different\n   formats suggested by their titles or descriptions. Deduplicate IDs and explain\n   the apparent differences without asserting that the actual content differs.\n   If results are unsuitable, refine the query once or ask for another link.\n3. Provide each selected title, channel, returned `watchUrl`, and metadata-based\n   reason for inclusion. Call `display-video({ videoId, reason? })` for the first\n   selection when viewing is wanted. Extract IDs from supplied YouTube URLs.\n   Accept only `status: ready` with a video; handle `invalid_video_id`,\n   `video_not_found`, or a tool error by explaining the issue and offering a\n   returned alternative. Do not render every candidate as a browsing gallery.\n4. Invite the user to share takeaways against the checklist after each video.\n   Open the next selection individually when requested. Wait for the user's\n   account instead of assuming playback, understanding, or specific claims.\n5. Build a comparison matrix from the supplied takeaways or pasted excerpts.\n   Attribute each observation to its source, mark missing dimensions unknown,\n   and distinguish a shared idea from a difference in presentation or a reported\n   disagreement. Ask for the relevant excerpt before interpreting an ambiguous\n   disagreement.\n6. Recommend which explanation appears more useful for this user's goal based\n   on their reported experience. List any unresolved question and one focused\n   next step; do not manufacture a winner when the evidence is incomplete.\n\n## Deliverable\n\nReturn a two-video selection and shared viewing checklist, followed by a\ncomparison matrix and a goal-specific recommendation once the user provides\ncontent. Keep selection evidence and user-reported content clearly attributed.\n\n## Boundaries\n\n- Do not require annotations, access private widget state, or claim to watch or\n  transcribe either video. A title-only discussion handoff supplies no takeaways.\n- Do not treat channel variety, popularity, or a title as proof of accuracy,\n  opposing viewpoints, independent sources, or teaching quality.\n- Leave reported factual disagreements unresolved unless supporting evidence\n  is available. Distinguish comparing explanations from verifying their claims.\n- Use only the registered tool arguments. Keep the matrix in chat; do not\n  promise synchronized players, automatic seeking, or a saved comparison board.\n- If the user explicitly requests text only or no video display, call neither\n  tool and compare only the material already provided.\n"
}