[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CB3CYCSAQENCW9SB58BYHC":3,"skill-same:01M7CB3CYCSAQENCW9SB58BYHC":98},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":70,"commitSha":71,"skillDirectory":72,"forksCount":73,"repositoryHtmlUrl":74,"repositoryHomepageUrl":75,"ownerHtmlUrl":76,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":77,"install":78,"resources":91},"01M7CB3CYCSAQENCW9SB58BYHC","short-drama-workflow",null,"Workflow for producing plot-dense short dramas with character and location reference sheets, complete 15-second Seedance segments, native dialogue, parallel multi-reference T2V generation, duration validation, and final assembly. Use for 爽文短剧、都市反转短剧、对白短剧、连续多段短剧，或用户明确调用 $short-drama-workflow. Do not use for product-centered scenario advertisements.","VideoVerses\u002FCuti-Harness","VideoVerses","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F175584708?v=4","01M7CAV1DS56HR4KMB99EX9KAH","01M7CAV1EC999NFVVG4K7N1VJS",74,"MIT","https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fshort-drama-workflow\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: short-drama-workflow\ndescription: >-\n  Workflow for producing plot-dense short dramas with character and location\n  reference sheets, complete 15-second Seedance segments, native dialogue,\n  parallel multi-reference T2V generation, duration validation, and final assembly. Use for\n  爽文短剧、都市反转短剧、对白短剧、连续多段短剧，或用户明确调用\n  $short-drama-workflow. Do not use for product-centered scenario advertisements.\nmetadata:\n  kind: workflow\n  version: \"1.1.0\"\n  workflow:\n    title: Short drama production\n    mode: short_drama_workflow\n    planning:\n      mode: staged\n      checkpoints:\n        - id: story_ready\n          after_phase: story_intent\n          next_phase: visual_production\n          required_artifacts: [story_draft]\n          resolves: [characters, scenes, shots, audio]\n          instruction: Complete dialogue, shared references, and parallel shot specifications from the story draft.\n    entrypoints: [text, image, audio, video]\n    dependencies:\n      # Planning guidance only. Runtime leaf steps still carry no legacy\n      # Director Skill injection and remain owned by this Workflow compiler.\n      skills: [seedance2]\n    parameters:\n      shot_workflow_mode: seedance2_short_drama\n      content_category: short_drama\n      segment_duration_seconds: 15\n      segment_execution_mode: parallel\n      video_generation_mode: t2v\n      continuity_mode: shared_reference_images\n    pipeline:\n      - atomic.text.generate\n      - atomic.image.generate\n      - atomic.video.generate\n      - media.concat\n      - media.transcribe\n      - subtitle.compose\n      - media.subtitle_burn\n    requires_keyframe: false\n    allowed_capabilities:\n      - actions.suggest\n      - atomic.text.generate\n      - atomic.image.generate\n      - atomic.video.generate\n      - api.provider.generate\n      - api.ark_protocol.generate\n      - media.concat\n      - media.transcribe\n      - subtitle.compose\n      - media.subtitle_burn\n---\n\n# Short Drama Workflow\n\nCreate a compact narrative in complete 15-second video units. Use the installed\nlegacy `seedance2` Skill for Seedance prompt grammar and multi-reference mechanics.\nGenerate through WaveSpeed Seedance 2.5 unless the user explicitly selects a\ndifferent available model. Do not load unrelated prompt-only helper bundles.\n\n## Workflow identity\n\nCarry these fields on every stage task:\n\n```json\n{\n  \"workflow_mode\": \"short_drama_workflow\",\n  \"activated_workflow\": \"short-drama-workflow\",\n  \"shot_workflow_mode\": \"seedance2_short_drama\",\n  \"content_category\": \"short_drama\"\n}\n```\n\nAlso preserve the project `thread_id`, chosen aspect ratio, resolution, model,\noutput language, and spoken-language contract across all stages.\n\n## Main production flow\n\n### 1. Lock the story and production bible\n\nUse `atomic.text.generate` once to create the executable production blueprint.\nThis is a language-model stage: omit `model`, or use a supported `llm_model` only.\nNever set `model: seedance-2.5` on this task; that video model belongs exclusively\nto `atomic.video.generate` segment tasks.\nIt must contain:\n\n- the complete causal story and ending;\n- a compact character bible with stable appearance, clothing, temperament, and\n  relationships;\n- a location and lighting bible for recurring spaces;\n- one dramatic function for each 15-second segment;\n- a complete time-coded beat plan for every segment, with a new action, reaction,\n  revelation, or power shift roughly every 1-2 seconds;\n- native dialogue, ambience, sound effects, music or deliberate silence;\n- a narrative handoff describing what changes at the end of each segment and what\n  the next segment must continue, without requiring pixel-matched boundary frames;\n- one generation-ready Seedance prompt per segment.\n\nFor a default 60-second request, produce four 15-second segments. For another\nduration, use the smallest sensible number of complete 15-second segments and\ntell the user if the requested duration requires rounding or a shorter final edit.\nDo not silently replace 15-second units with 5- or 6-second clips.\n\n### 2. Generate character and location reference images\n\nBefore any video generation, create and persist the setting images described by\nthe approved production blueprint.\n\nUse `atomic.image.generate` to produce:\n\n- a character setting reference for every recurring principal character, showing\n  stable face, hair, apparent age, body proportions, costume, palette, and one or\n  more useful views; use an ensemble sheet only when it remains easy to identify\n  each person unambiguously;\n- a scene setting reference for every recurring primary location, showing spatial\n  layout, entrances, important furniture or props, lighting direction, time of day,\n  palette, and the camera-facing geography needed by the script.\n\nUse the configured image-generation tool selected for the run. Do not silently\nswitch providers. Keep these images neutral and reusable: they define identity and\nspace, not a frozen opening shot from one segment.\n\nPersist each image as a visible workspace artifact with a clear semantic title,\nsuch as `character_setting_reference` or `scene_setting_reference`, and record which\ncharacters or locations it covers. Do not begin video generation until all required\nreferences exist and are readable.\n\nThese are setting references, not per-shot keyframes. `requires_keyframe: false`\nstill applies: do not create a first-frame gallery or force every segment to begin\nwith the same composition.\n\n### 3. Build dense dramatic causality\n\nEvery beat must change the situation. Prefer a readable progression such as:\n\n1. immediate conflict, accusation, danger, or impossible situation;\n2. the protagonist absorbs pressure while preparing a countermove;\n3. evidence, leverage, identity, or intent is revealed in escalating steps;\n4. the opponent interrupts, denies, or overcommits;\n5. the power relationship reverses through an earned consequence;\n6. end on a restrained payoff or a new hook.\n\nThis is a decision pattern, not a fixed plot template. Adapt it to suspense,\nworkplace, family, romance, comedy, fantasy, or another user-selected genre.\nPreserve the user's premise and ending rather than forcing a workplace reveal.\n\n### 4. Compile each 15-second generation prompt\n\nEach prompt must be self-sufficient and include:\n\n- model, exact duration, aspect ratio, and resolution;\n- character and location anchors relevant to the current segment;\n- time-coded actions covering the whole 15 seconds;\n- short dialogue lines labeled by speaker and delivery;\n- stable shot sizes during spoken lines when lip synchronization matters;\n- native ambience and concrete sound cues;\n- the closing dramatic beat and the narrative state inherited by the next segment;\n- only the negative constraints justified by the brief.\n\nAttach the relevant character and scene setting images to every segment as reusable\nidentity and environment references. In the prompt, state what each reference owns:\ncharacter references control identity and costume; scene references control layout,\nlighting, and recurring props. They must not dictate the segment's first-frame\ncomposition or replace its time-coded action plan.\n\nKeep dialogue speakable within its assigned time. Prefer one speaker at a time\nand short turns over overlapping speeches. Do not use narration unless requested.\n\nFollow the run's language contract:\n\n- user-visible plans and artifacts use the selected output language;\n- dialogue and narration use the selected spoken language;\n- provider prompts may remain in Chinese when that helps legacy `seedance2`, but\n  they must explicitly state the required spoken language and preserve dialogue\n  verbatim;\n- words such as “英文短剧”, “English-language video”, or “全片英语” count as an\n  explicit English spoken-language request even if the surrounding request is Chinese;\n- never translate explicit dialogue into the UI language.\n\nPlan every segment boundary as an intentional editorial cut, reaction cut, location\nchange, insert, or motivated time jump. Parallel T2V generation does not guarantee\nthat one segment's first frame is pixel-identical to another segment's last frame.\n\n### 5. Generate all segments in parallel with multi-reference T2V\n\nAfter the blueprint and all required character and scene references are ready,\nschedule every segment as an independent `atomic.video.generate` task. Segment tasks\nmay run in parallel; none may depend on another segment task. Each depends only on\nthe approved blueprint and the setting artifacts relevant to that segment.\n\nUse:\n\n- `provider: wavespeed` when provider selection is exposed;\n- `model: seedance-2.5` or the capability's exact equivalent;\n- `generation_mode: \"t2v\"`;\n- `duration: 15` and `duration_seconds: 15`;\n- the run's resolution and aspect ratio;\n- native dialogue and ambience.\n\nPass the relevant setting artifacts through the capability's multi-reference image\nfield (`reference_images` or its exact supported equivalent). Use reference-driven\ntext-to-video behavior rather than treating the first character or scene sheet as a\nforced first frame.\n\nEvery segment must receive all references needed for the identities and location it\ncontains. Do not attach irrelevant references merely to maximize image count. Bind\neach supplied image to its role in the prompt, for example: character identity and\ncostume, recurring location layout and lighting, or a persistent hero prop.\n\nDo not set `start_image_url`, `end_image_url`, or `generation_mode: \"i2v\"`. Do not\nextract or pass a previous segment's tail frame. Those choices would serialize the\ngraph and prevent the multi-reference T2V behavior required by this workflow.\n\nVerify the returned media duration before using it. A nominal request is not\nproof that the provider returned 15 seconds. If the segment is materially short,\nreject it, strengthen the explicit duration contract, and retry that segment once.\nOne segment's failure must not cause successful sibling segments to be regenerated.\n\n### 6. Assemble the approved segments\n\nCall `media.concat` with the accepted videos in narrative order and\n`normalize: true`.\n\n- Use `transition_duration: 0` by default because boundaries were designed as\n  intentional cuts. Add a transition only when the script explicitly calls for it.\n- Check assembled duration, ordering, audio presence, and each join.\n- Persist the assembled MP4 as the visible final workspace artifact; do not return\n  only a raw URL in chat.\n\n## Optional captions\n\nCaptions are not part of the default generation path. Add them only when the user\nasks for subtitles or approves them in the original brief:\n\n1. transcribe the assembled video with `media.transcribe` using the spoken language;\n2. create audio-timed subtitles with `subtitle.compose`;\n3. validate timestamps, overlaps, and reading speed;\n4. burn them with `media.subtitle_burn` using a bottom-safe, language-compatible\n   font and a restrained size;\n5. persist the captioned MP4 as a new visible artifact without replacing the clean\n   master.\n\n## Acceptance gates\n\nDo not advance past a segment when any of these are false:\n\n- all required character and scene setting artifacts exist before segment 1;\n- every segment received the relevant setting references;\n- every segment used multi-reference T2V rather than I2V or tail-frame chaining;\n- its actual duration satisfies the intended 15-second unit;\n- required dialogue language matches the run contract;\n- the segment performs its assigned dramatic function;\n- identity, costume, location, lighting, and recurring props agree with the shared\n  references, while boundary frames are evaluated as intentional cuts rather than\n  pixel-equal continuations;\n- the output contains a usable video artifact and audio when native audio was requested.\n\nRetry a failed generation at most once unless the user explicitly asks for further\nattempts. Provider credit, authentication, or unsupported-model errors are blockers,\nnot creative failures; report them immediately instead of repeatedly resubmitting.\n","# Short Drama Workflow\n\nCreate a compact narrative in complete 15-second video units. Use the installed\nlegacy `seedance2` Skill for Seedance prompt grammar and multi-reference mechanics.\nGenerate through WaveSpeed Seedance 2.5 unless the user explicitly selects a\ndifferent available model. Do not load unrelated prompt-only helper bundles.\n\n## Workflow identity\n\nCarry these fields on every stage task:\n\n```json\n{\n  \"workflow_mode\": \"short_drama_workflow\",\n  \"activated_workflow\": \"short-drama-workflow\",\n  \"shot_workflow_mode\": \"seedance2_short_drama\",\n  \"content_category\": \"short_drama\"\n}\n```\n\nAlso preserve the project `thread_id`, chosen aspect ratio, resolution, model,\noutput language, and spoken-language contract across all stages.\n\n## Main production flow\n\n### 1. Lock the story and production bible\n\nUse `atomic.text.generate` once to create the executable production blueprint.\nThis is a language-model stage: omit `model`, or use a supported `llm_model` only.\nNever set `model: seedance-2.5` on this task; that video model belongs exclusively\nto `atomic.video.generate` segment tasks.\nIt must contain:\n\n- the complete causal story and ending;\n- a compact character bible with stable appearance, clothing, temperament, and\n  relationships;\n- a location and lighting bible for recurring spaces;\n- one dramatic function for each 15-second segment;\n- a complete time-coded beat plan for every segment, with a new action, reaction,\n  revelation, or power shift roughly every 1-2 seconds;\n- native dialogue, ambience, sound effects, music or deliberate silence;\n- a narrative handoff describing what changes at the end of each segment and what\n  the next segment must continue, without requiring pixel-matched boundary frames;\n- one generation-ready Seedance prompt per segment.\n\nFor a default 60-second request, produce four 15-second segments. For another\nduration, use the smallest sensible number of complete 15-second segments and\ntell the user if the requested duration requires rounding or a shorter final edit.\nDo not silently replace 15-second units with 5- or 6-second clips.\n\n### 2. Generate character and location reference images\n\nBefore any video generation, create and persist the setting images described by\nthe approved production blueprint.\n\nUse `atomic.image.generate` to produce:\n\n- a character setting reference for every recurring principal character, showing\n  stable face, hair, apparent age, body proportions, costume, palette, and one or\n  more useful views; use an ensemble sheet only when it remains easy to identify\n  each person unambiguously;\n- a scene setting reference for every recurring primary location, showing spatial\n  layout, entrances, important furniture or props, lighting direction, time of day,\n  palette, and the camera-facing geography needed by the script.\n\nUse the configured image-generation tool selected for the run. Do not silently\nswitch providers. Keep these images neutral and reusable: they define identity and\nspace, not a frozen opening shot from one segment.\n\nPersist each image as a visible workspace artifact with a clear semantic title,\nsuch as `character_setting_reference` or `scene_setting_reference`, and record which\ncharacters or locations it covers. Do not begin video generation until all required\nreferences exist and are readable.\n\nThese are setting references, not per-shot keyframes. `requires_keyframe: false`\nstill applies: do not create a first-frame gallery or force every segment to begin\nwith the same composition.\n\n### 3. Build dense dramatic causality\n\nEvery beat must change the situation. Prefer a readable progression such as:\n\n1. immediate conflict, accusation, danger, or impossible situation;\n2. the protagonist absorbs pressure while preparing a countermove;\n3. evidence, leverage, identity, or intent is revealed in escalating steps;\n4. the opponent interrupts, denies, or overcommits;\n5. the power relationship reverses through an earned consequence;\n6. end on a restrained payoff or a new hook.\n\nThis is a decision pattern, not a fixed plot template. Adapt it to suspense,\nworkplace, family, romance, comedy, fantasy, or another user-selected genre.\nPreserve the user's premise and ending rather than forcing a workplace reveal.\n\n### 4. Compile each 15-second generation prompt\n\nEach prompt must be self-sufficient and include:\n\n- model, exact duration, aspect ratio, and resolution;\n- character and location anchors relevant to the current segment;\n- time-coded actions covering the whole 15 seconds;\n- short dialogue lines labeled by speaker and delivery;\n- stable shot sizes during spoken lines when lip synchronization matters;\n- native ambience and concrete sound cues;\n- the closing dramatic beat and the narrative state inherited by the next segment;\n- only the negative constraints justified by the brief.\n\nAttach the relevant character and scene setting images to every segment as reusable\nidentity and environment references. In the prompt, state what each reference owns:\ncharacter references control identity and costume; scene references control layout,\nlighting, and recurring props. They must not dictate the segment's first-frame\ncomposition or replace its time-coded action plan.\n\nKeep dialogue speakable within its assigned time. Prefer one speaker at a time\nand short turns over overlapping speeches. Do not use narration unless requested.\n\nFollow the run's language contract:\n\n- user-visible plans and artifacts use the selected output language;\n- dialogue and narration use the selected spoken language;\n- provider prompts may remain in Chinese when that helps legacy `seedance2`, but\n  they must explicitly state the required spoken language and preserve dialogue\n  verbatim;\n- words such as “英文短剧”, “English-language video”, or “全片英语” count as an\n  explicit English spoken-language request even if the surrounding request is Chinese;\n- never translate explicit dialogue into the UI language.\n\nPlan every segment boundary as an intentional editorial cut, reaction cut, location\nchange, insert, or motivated time jump. Parallel T2V generation does not guarantee\nthat one segment's first frame is pixel-identical to another segment's last frame.\n\n### 5. Generate all segments in parallel with multi-reference T2V\n\nAfter the blueprint and all required character and scene references are ready,\nschedule every segment as an independent `atomic.video.generate` task. Segment tasks\nmay run in parallel; none may depend on another segment task. Each depends only on\nthe approved blueprint and the setting artifacts relevant to that segment.\n\nUse:\n\n- `provider: wavespeed` when provider selection is exposed;\n- `model: seedance-2.5` or the capability's exact equivalent;\n- `generation_mode: \"t2v\"`;\n- `duration: 15` and `duration_seconds: 15`;\n- the run's resolution and aspect ratio;\n- native dialogue and ambience.\n\nPass the relevant setting artifacts through the capability's multi-reference image\nfield (`reference_images` or its exact supported equivalent). Use reference-driven\ntext-to-video behavior rather than treating the first character or scene sheet as a\nforced first frame.\n\nEvery segment must receive all references needed for the identities and location it\ncontains. Do not attach irrelevant references merely to maximize image count. Bind\neach supplied image to its role in the prompt, for example: character identity and\ncostume, recurring location layout and lighting, or a persistent hero prop.\n\nDo not set `start_image_url`, `end_image_url`, or `generation_mode: \"i2v\"`. Do not\nextract or pass a previous segment's tail frame. Those choices would serialize the\ngraph and prevent the multi-reference T2V behavior required by this workflow.\n\nVerify the returned media duration before using it. A nominal request is not\nproof that the provider returned 15 seconds. If the segment is materially short,\nreject it, strengthen the explicit duration contract, and retry that segment once.\nOne segment's failure must not cause successful sibling segments to be regenerated.\n\n### 6. Assemble the approved segments\n\nCall `media.concat` with the accepted videos in narrative order and\n`normalize: true`.\n\n- Use `transition_duration: 0` by default because boundaries were designed as\n  intentional cuts. Add a transition only when the script explicitly calls for it.\n- Check assembled duration, ordering, audio presence, and each join.\n- Persist the assembled MP4 as the visible final workspace artifact; do not return\n  only a raw URL in chat.\n\n## Optional captions\n\nCaptions are not part of the default generation path. Add them only when the user\nasks for subtitles or approves them in the original brief:\n\n1. transcribe the assembled video with `media.transcribe` using the spoken language;\n2. create audio-timed subtitles with `subtitle.compose`;\n3. validate timestamps, overlaps, and reading speed;\n4. burn them with `media.subtitle_burn` using a bottom-safe, language-compatible\n   font and a restrained size;\n5. persist the captioned MP4 as a new visible artifact without replacing the clean\n   master.\n\n## Acceptance gates\n\nDo not advance past a segment when any of these are false:\n\n- all required character and scene setting artifacts exist before segment 1;\n- every segment received the relevant setting references;\n- every segment used multi-reference T2V rather than I2V or tail-frame chaining;\n- its actual duration satisfies the intended 15-second unit;\n- required dialogue language matches the run contract;\n- the segment performs its assigned dramatic function;\n- identity, costume, location, lighting, and recurring props agree with the shared\n  references, while boundary frames are evaluated as intentional cuts rather than\n  pixel-equal continuations;\n- the output contains a usable video artifact and audio when native audio was requested.\n\nRetry a failed generation at most once unless the user explicitly asks for further\nattempts. Provider credit, authentication, or unsupported-model errors are blockers,\nnot creative failures; report them immediately instead of repeatedly resubmitting.\n",{"name":5,"metadata":23,"description":7},{"kind":24,"version":25,"workflow":26},"workflow","1.1.0",{"mode":27,"title":28,"pipeline":29,"planning":37,"parameters":52,"entrypoints":59,"dependencies":63,"requires_keyframe":19,"allowed_capabilities":66},"short_drama_workflow","Short drama production",[30,31,32,33,34,35,36],"atomic.text.generate","atomic.image.generate","atomic.video.generate","media.concat","media.transcribe","subtitle.compose","media.subtitle_burn",{"mode":38,"checkpoints":39},"staged",[40],{"id":41,"resolves":42,"next_phase":47,"after_phase":48,"instruction":49,"required_artifacts":50},"story_ready",[43,44,45,46],"characters","scenes","shots","audio","visual_production","story_intent","Complete dialogue, shared references, and parallel shot specifications from the story draft.",[51],"story_draft",{"continuity_mode":53,"content_category":54,"shot_workflow_mode":55,"video_generation_mode":56,"segment_execution_mode":57,"segment_duration_seconds":58},"shared_reference_images","short_drama","seedance2_short_drama","t2v","parallel",15,[60,61,46,62],"text","image","video",{"skills":64},[65],"seedance2",[67,30,31,32,68,69,33,34,35,36],"actions.suggest","api.provider.generate","api.ark_protocol.generate","main","60364690293feed5d9b458e19a441d4d7561e65e","services\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fshort-drama-workflow",1,"https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness","https:\u002F\u002Fnewai.land\u002Fcuti-harness","https:\u002F\u002Fgithub.com\u002FVideoVerses","2026-10-11 19:27:56",{"source":8,"skillName":5,"commands":79,"prompt":89,"zipAvailable":90},[80,83,86],{"tool":81,"command":82},"npx","npx skills add VideoVerses\u002FCuti-Harness --skill short-drama-workflow",{"tool":84,"command":85},"bunx","bunx skills add VideoVerses\u002FCuti-Harness --skill short-drama-workflow",{"tool":87,"command":88},"pnpm","pnpm dlx skills add VideoVerses\u002FCuti-Harness --skill short-drama-workflow","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fshort-drama-workflow\u002FSKILL.md",true,[92],{"path":93,"resourceType":94,"resourceTypeDesc":95,"sizeBytes":96,"sourceHtmlUrl":97},"agents\u002Fopenai.yaml","REFERENCE","参考文档",332,"https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fshort-drama-workflow\u002Fagents\u002Fopenai.yaml",[99,106,113,120,127,134],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":18},"01M7CB3T5C44KCS1RRXMXXNCHF","external-shot-continuity","External continuity review method for long-form video projects.","https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Ftests\u002Ffixtures\u002Fexternal-skills\u002Fexternal-shot-continuity\u002FSKILL.md",[],[],{"skillNo":107,"name":108,"nameZh":6,"description":109,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":18},"01M7CB3PKKR1JRB3XAG5S9T7NS","independent-story-workflow","Create a three-part story document without paid media generation.","https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Ftests\u002Ffixtures\u002Fdynamic-workflow\u002FSKILL.md",[],[],{"skillNo":114,"name":115,"nameZh":6,"description":116,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":117,"categories":118,"keywords":119,"publishedAt":18},"01M7CB3K7RMRQ4WDYAYNC1S1S0","suno-song","Instruction helper: write a singable original Suno song from Bitwize (verbatim copy), then call suno.generate. Trigger: 原创歌曲, 写歌, Suno lyrics, $suno-song. Not a workflow.","https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fsuno-song\u002FSKILL.md",[],[],{"skillNo":121,"name":65,"nameZh":6,"description":122,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":123,"sourceHtmlUrl":124,"categories":125,"keywords":126,"publishedAt":18},"01M7CB36GV8A2YX4P41457FRZ6","Workflow: 即梦 Seedance 视频创意工作台。图+文案→看图分析→文案扩写→运镜→API生成。 触发词：即梦、Seedance、seedance、视频提示词、AI视频、运镜、短剧广告、$seedance2.","https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fseedance2\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fseedance2\u002FSKILL.md",[],[],{"skillNo":128,"name":129,"nameZh":6,"description":130,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":131,"categories":132,"keywords":133,"publishedAt":18},"01M7CB2ZDR784PMJ60NZ616GWQ","product-voiceover-narration","Design concise native voiceover for commercial product-video segments and bind each spoken line to verified selling points, visible actions, and exact timeline ranges. Use when a product advertisement should introduce the product through narration, when Seedance 2.0 should generate picture and voice together, or when product shots need coordinated narration, music, and sound direction. Do not synthesize a separate voice track and never invent product claims.","https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fproduct-voiceover-narration\u002FSKILL.md",[],[],{"skillNo":135,"name":136,"nameZh":6,"description":137,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":138,"categories":139,"keywords":140,"publishedAt":18},"01M7CB2QZZZPKS0P8ECQ5QX16P","product-feature-demo-script","Turn verified product features and selling points into visual commercial shot scripts with a clear trigger, product response, and visible user-facing result. Use for product advertisements that need functional demonstrations, interaction, before-and-after change, environmental stress, usage scenarios, or conceptual mechanism visualization instead of repeated rotations, orbit shots, floating packshots, and exterior-only beauty shots. Suitable for phones, cameras, headphones, watches, computers, appliances, tools, cosmetics devices, and other functional products. Never invent unsupported capabilities or performance claims.","https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FCuti-Harness\u002Fblob\u002Fmain\u002Fservices\u002Fvideo-runtime\u002Fskills\u002Fexternal\u002Fproduct-feature-demo-script\u002FSKILL.md",[],[]]