[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6BXZVGARMHGZSX3XQA3RK8D":3,"skill-same:01M6BXZVGARMHGZSX3XQA3RK8D":74},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":23,"publishedAt":46,"linkOnly":47,"markdownContent":48,"markdownBody":49,"frontMatter":50,"refName":51,"commitSha":52,"skillDirectory":53,"forksCount":54,"repositoryHtmlUrl":55,"repositoryHomepageUrl":56,"ownerHtmlUrl":57,"ownerBlogUrl":12,"ownerDescription":58,"updatedAt":59,"install":60,"resources":73},"01M6BXZVGARMHGZSX3XQA3RK8D","gpt-image-2-prompting",null,"Prompt OpenAI's GPT Image 2 — the five-slot Scene\u002FSubject\u002FDetails\u002FUse case\u002FConstraints template it responds to, the change-versus-preserve shape for edits, labelled multi-image compositing, and how to get text on an image to render verbatim. Use whenever the model id contains gpt-image-2 (openai\u002Fgpt-image-2, openai\u002Fgpt-image-2\u002Fedit, openai\u002Fgpt-image-2\u002Ftext-to-image, kie's gpt-image-2-text-to-image and gpt-image-2-image-to-image) on generate_image, edit_image or a TextToImage node. Not for GPT Image 1, 1-mini, 1.5 or 2.5.","nodetool-ai\u002Fnodetool","nodetool-ai","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F164873569?v=4","01M6BXPFDW36K9K1MR8ZH63HG2","nodetool.ai","01M6BXPFJ893XM219GQN60B0WK",635,"AGPL-3.0","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fgpt-image-2-prompting\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000016","图像与多媒体","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45],"gpt","prompting","gpt image 2","image","gpt image 2 prompting","kie","for","render","how to","get","use","edit","node","model","text","template","how","constraints","Text-to-Image","openai","generate","change","2026-11-10 10:12:30",false,"---\nname: gpt-image-2-prompting\ndescription: Prompt OpenAI's GPT Image 2 — the five-slot Scene\u002FSubject\u002FDetails\u002FUse case\u002FConstraints template it responds to, the change-versus-preserve shape for edits, labelled multi-image compositing, and how to get text on an image to render verbatim. Use whenever the model id contains gpt-image-2 (openai\u002Fgpt-image-2, openai\u002Fgpt-image-2\u002Fedit, openai\u002Fgpt-image-2\u002Ftext-to-image, kie's gpt-image-2-text-to-image and gpt-image-2-image-to-image) on generate_image, edit_image or a TextToImage node. Not for GPT Image 1, 1-mini, 1.5 or 2.5.\n---\n\n# GPT Image 2 → five slots, in order\n\nThe model responds to structure: scene, subject, specific details, intended\nartifact, constraints, in that order, with line breaks between sections once\nthe prompt runs past a short paragraph. Everything else sits on top of that\nspine.\n\nReach it with `find_model` for `text_to_image` or `image_to_image`, then\n`generate_image` \u002F `edit_image`. `generate_image` forwards only the prompt, a\nnegative prompt, the size and a quality string, and `edit_image` only the\nimages, the prompt, a negative prompt, a target size and a strength. Output\nformat and `background` exist only on the route's own node: find it with\n`search_nodes`, read its fields with `get_node_info`, and run it with\n`run_node`.\n\n## The template\n\n```\nScene:\n[where this happens, time of day, background, environment]\n\nSubject:\n[who or what is the main focus]\n\nImportant details:\n[materials, clothing, texture, lighting, camera angle, lens feel, composition, mood]\n\nUse case:\n[editorial photo \u002F product mockup \u002F poster \u002F UI screen \u002F infographic \u002F concept frame]\n\nConstraints:\n[no watermark \u002F no logos \u002F no extra text \u002F preserve face \u002F preserve layout]\n```\n\nFive slots, five problems people usually blur together: where the image exists,\nwhat it is about, what must be visible, what kind of finished artifact you\nwant, and what must not drift. The fifth is where mediocre prompts fail\nsilently — describe the idea without bounding it and the model gets inventive\nin directions you will regret.\n\nFilled in:\n\n```\nScene:\nA quiet classical museum gallery in soft afternoon light.\n\nSubject:\nA woman in her 30s standing casually in front of a large oil painting.\n\nImportant details:\nNatural smile, realistic skin texture, beige knit sweater, dark jeans, white\nsneakers, eye-level full-body framing, marble floor reflections, warm neutral\ncolour balance, shallow depth of field, believable indoor ambient light.\n\nUse case:\nEditorial lifestyle photograph.\n\nConstraints:\nNo watermark, no logos, no extra people in the foreground, no heavy retouching.\n```\n\n## Rules\n\n1. **Visual facts over vague praise.** Drop stunning, incredible, epic,\n   masterpiece, insane detail. Use overcast daylight, brushed aluminium,\n   chipped paint, clean kerning, 50mm feel, soft bounce light, worn canvas.\n2. **Style tags need visual targets.** \"Minimalist brutalist editorial luxury\n   premium\" is noise. \"Cream background, heavy black condensed sans serif,\n   asymmetrical type block, one hero object, generous negative space, studio\n   tabletop lighting\" is a layout.\n3. **Say the real thing.** Transit kiosk. Boarding pass. Preserve the face.\n   Mood language buries the brief.\n4. **Separate change from preserve in edits.** \"Change only X\", \"keep\n   everything else the same\", and repeat the preserve list on every pass.\n5. **Treat text as typography.** Wrap literal copy in quotes or ALL CAPS, mark\n   it EXACT TEXT, and specify font class, size, colour and placement. Add \"no\n   extra words\" and \"no duplicate text\".\n6. **One revision per turn.** Small iterative edits beat one giant rewrite.\n\n## Three modes\n\n**Generate from scratch** — the five-slot template above. One clean pass lands\nbelievable mundane realism once the prompt locks the lighting, the camera\nbehaviour and the environment details.\n\n**Edit one image** — two columns, plus a physical-realism line:\n\n```\nChange:\nReplace the parked car with a vintage bicycle.\n\nPreserve:\nThe house, fence, driveway concrete, landscaping, lighting direction, and time\nof day, exactly.\n\nConstraints:\nMatch the bicycle's scale and shadow pattern to the existing scene. No extra\nobjects, no redesign, no watermark.\n```\n\nThe preserve list carries the edit. Inventory what must stay — awning, brick\nfacade, mullions, reflections, sidewalk, every person on it — and the edit\nstays in scope.\n\n**Combine multiple images** — label every input by role and reference the\nlabels in the instruction. fal's edit route and kie's image-to-image route take\nup to 16 reference images; the other routes state no cap.\n\n```\nImage 1: base scene to preserve.\nImage 2: jacket reference.\nImage 3: boots reference.\n\nDress the person from Image 1 using the jacket from Image 2 and the boots from\nImage 3. Preserve the face, body shape, pose, background, lighting, and framing\nfrom Image 1. Fit the garments naturally with realistic folds, drape, occlusion\nand contact shadows. No extra accessories.\n```\n\nUnlabelled inputs make the model guess which image is content and which is\nreference, and it guesses wrong.\n\n## What it is good at\n\n- **Photoreal editorial.** Describe the photograph, not the fantasy: lens,\n  framing, time of day, light source, surface wear, an ordinary background\n  detail, one believable imperfection.\n- **Product.** Material accuracy, lighting consistency, label fidelity. A flat\n  inventory of physical objects plus one piece of print that must stay legible\n  is the entire recipe.\n- **UI and screenshots.** Name the screen type, the hierarchy, the exact copy,\n  the state, and the layout logic. \"Clean survival HUD along the bottom,\n  believable UI spacing\" does the layout work; remove those clauses and the HUD\n  collapses into noise.\n- **Text in image.** Give the copy verbatim, then typography, then layout, then\n  \"render the text verbatim \u002F no extra words \u002F no duplicate text\". For a still\n  that will sit on a timeline or seed a video clip, leave the copy off and add\n  it as a text clip afterwards (`caption-titles`) so it stays editable.\n- **Style transfer.** \"Same style\" is not enough. Name the parts: chunky pixel\n  forms, limited arcade palette, bright glow accents, clean silhouette edges.\n- **Drawing to photo.** Say whether the drawing is a suggestion or a contract:\n  \"preserve the exact layout, horizon line, proportions, river path, mountain\n  placement, tree placement, and overall perspective\".\n- **Character consistency.** First prompt establishes the anchor. Second prompt\n  repeats the anchor details verbatim and adds \"do not redesign the character\".\n\nTransparency needs a route with a `background` field set to `transparent`:\nkie's (only at 1K resolution, per its field note) or Replicate's\n`openai\u002Fgpt-image-2`, with PNG or WebP output. The fal and AtlasCloud routes\nhave no `background` field.\n\nCheck the render rather than assuming it: `critique_image`, or\n`score_image_adherence` when the prompt has copy or a layout to hold.\n\nAdapted from fal's GPT Image 2 prompting guide:\nhttps:\u002F\u002Ffal.ai\u002Flearn\u002Ftools\u002Fprompting-gpt-image-2\n","# GPT Image 2 → five slots, in order\n\nThe model responds to structure: scene, subject, specific details, intended\nartifact, constraints, in that order, with line breaks between sections once\nthe prompt runs past a short paragraph. Everything else sits on top of that\nspine.\n\nReach it with `find_model` for `text_to_image` or `image_to_image`, then\n`generate_image` \u002F `edit_image`. `generate_image` forwards only the prompt, a\nnegative prompt, the size and a quality string, and `edit_image` only the\nimages, the prompt, a negative prompt, a target size and a strength. Output\nformat and `background` exist only on the route's own node: find it with\n`search_nodes`, read its fields with `get_node_info`, and run it with\n`run_node`.\n\n## The template\n\n```\nScene:\n[where this happens, time of day, background, environment]\n\nSubject:\n[who or what is the main focus]\n\nImportant details:\n[materials, clothing, texture, lighting, camera angle, lens feel, composition, mood]\n\nUse case:\n[editorial photo \u002F product mockup \u002F poster \u002F UI screen \u002F infographic \u002F concept frame]\n\nConstraints:\n[no watermark \u002F no logos \u002F no extra text \u002F preserve face \u002F preserve layout]\n```\n\nFive slots, five problems people usually blur together: where the image exists,\nwhat it is about, what must be visible, what kind of finished artifact you\nwant, and what must not drift. The fifth is where mediocre prompts fail\nsilently — describe the idea without bounding it and the model gets inventive\nin directions you will regret.\n\nFilled in:\n\n```\nScene:\nA quiet classical museum gallery in soft afternoon light.\n\nSubject:\nA woman in her 30s standing casually in front of a large oil painting.\n\nImportant details:\nNatural smile, realistic skin texture, beige knit sweater, dark jeans, white\nsneakers, eye-level full-body framing, marble floor reflections, warm neutral\ncolour balance, shallow depth of field, believable indoor ambient light.\n\nUse case:\nEditorial lifestyle photograph.\n\nConstraints:\nNo watermark, no logos, no extra people in the foreground, no heavy retouching.\n```\n\n## Rules\n\n1. **Visual facts over vague praise.** Drop stunning, incredible, epic,\n   masterpiece, insane detail. Use overcast daylight, brushed aluminium,\n   chipped paint, clean kerning, 50mm feel, soft bounce light, worn canvas.\n2. **Style tags need visual targets.** \"Minimalist brutalist editorial luxury\n   premium\" is noise. \"Cream background, heavy black condensed sans serif,\n   asymmetrical type block, one hero object, generous negative space, studio\n   tabletop lighting\" is a layout.\n3. **Say the real thing.** Transit kiosk. Boarding pass. Preserve the face.\n   Mood language buries the brief.\n4. **Separate change from preserve in edits.** \"Change only X\", \"keep\n   everything else the same\", and repeat the preserve list on every pass.\n5. **Treat text as typography.** Wrap literal copy in quotes or ALL CAPS, mark\n   it EXACT TEXT, and specify font class, size, colour and placement. Add \"no\n   extra words\" and \"no duplicate text\".\n6. **One revision per turn.** Small iterative edits beat one giant rewrite.\n\n## Three modes\n\n**Generate from scratch** — the five-slot template above. One clean pass lands\nbelievable mundane realism once the prompt locks the lighting, the camera\nbehaviour and the environment details.\n\n**Edit one image** — two columns, plus a physical-realism line:\n\n```\nChange:\nReplace the parked car with a vintage bicycle.\n\nPreserve:\nThe house, fence, driveway concrete, landscaping, lighting direction, and time\nof day, exactly.\n\nConstraints:\nMatch the bicycle's scale and shadow pattern to the existing scene. No extra\nobjects, no redesign, no watermark.\n```\n\nThe preserve list carries the edit. Inventory what must stay — awning, brick\nfacade, mullions, reflections, sidewalk, every person on it — and the edit\nstays in scope.\n\n**Combine multiple images** — label every input by role and reference the\nlabels in the instruction. fal's edit route and kie's image-to-image route take\nup to 16 reference images; the other routes state no cap.\n\n```\nImage 1: base scene to preserve.\nImage 2: jacket reference.\nImage 3: boots reference.\n\nDress the person from Image 1 using the jacket from Image 2 and the boots from\nImage 3. Preserve the face, body shape, pose, background, lighting, and framing\nfrom Image 1. Fit the garments naturally with realistic folds, drape, occlusion\nand contact shadows. No extra accessories.\n```\n\nUnlabelled inputs make the model guess which image is content and which is\nreference, and it guesses wrong.\n\n## What it is good at\n\n- **Photoreal editorial.** Describe the photograph, not the fantasy: lens,\n  framing, time of day, light source, surface wear, an ordinary background\n  detail, one believable imperfection.\n- **Product.** Material accuracy, lighting consistency, label fidelity. A flat\n  inventory of physical objects plus one piece of print that must stay legible\n  is the entire recipe.\n- **UI and screenshots.** Name the screen type, the hierarchy, the exact copy,\n  the state, and the layout logic. \"Clean survival HUD along the bottom,\n  believable UI spacing\" does the layout work; remove those clauses and the HUD\n  collapses into noise.\n- **Text in image.** Give the copy verbatim, then typography, then layout, then\n  \"render the text verbatim \u002F no extra words \u002F no duplicate text\". For a still\n  that will sit on a timeline or seed a video clip, leave the copy off and add\n  it as a text clip afterwards (`caption-titles`) so it stays editable.\n- **Style transfer.** \"Same style\" is not enough. Name the parts: chunky pixel\n  forms, limited arcade palette, bright glow accents, clean silhouette edges.\n- **Drawing to photo.** Say whether the drawing is a suggestion or a contract:\n  \"preserve the exact layout, horizon line, proportions, river path, mountain\n  placement, tree placement, and overall perspective\".\n- **Character consistency.** First prompt establishes the anchor. Second prompt\n  repeats the anchor details verbatim and adds \"do not redesign the character\".\n\nTransparency needs a route with a `background` field set to `transparent`:\nkie's (only at 1K resolution, per its field note) or Replicate's\n`openai\u002Fgpt-image-2`, with PNG or WebP output. The fal and AtlasCloud routes\nhave no `background` field.\n\nCheck the render rather than assuming it: `critique_image`, or\n`score_image_adherence` when the prompt has copy or a layout to hold.\n\nAdapted from fal's GPT Image 2 prompting guide:\nhttps:\u002F\u002Ffal.ai\u002Flearn\u002Ftools\u002Fprompting-gpt-image-2\n",{"name":5,"description":7},"main","c7226baa82daca42a19a5855c116544f97aa8bac","packages\u002Fsystem-skills\u002Fgpt-image-2-prompting",91,"https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool","https:\u002F\u002Fnodetool.ai\u002F","https:\u002F\u002Fgithub.com\u002Fnodetool-ai","No-code AI development","2026-11-15 03:27:49",{"source":8,"skillName":5,"commands":61,"prompt":71,"zipAvailable":72},[62,65,68],{"tool":63,"command":64},"npx","npx skills add nodetool-ai\u002Fnodetool --skill gpt-image-2-prompting",{"tool":66,"command":67},"bunx","bunx skills add nodetool-ai\u002Fnodetool --skill gpt-image-2-prompting",{"tool":69,"command":70},"pnpm","pnpm dlx skills add nodetool-ai\u002Fnodetool --skill gpt-image-2-prompting","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fgpt-image-2-prompting\u002FSKILL.md",true,[],[75,84,117,140,155,171],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":79,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":83},"01M742MX3Y2ZYK884ZXTRFKB9P","humanizer","Rewrite AI-sounding text so it reads like the writer without changing what it says.\nUse when editing or reviewing prose for AI tells: not-X-but-Y contrasts, one-line\nclosers, staged openers, forced triads, dashes everywhere, inflated claims, sales\nlanguage, stock AI words, bold labels, or filler.","MIT","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fhumanizer\u002FSKILL.md",[],[],"2026-11-11 15:11:14",{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":98,"publishedAt":46},"01M6BY4D93R0Q9400ZHJ04DXP1","nodetool","Use the NodeTool MCP server to build, validate, run, and debug visual AI workflows, and to work with assets, collections, and jobs. Use when the user mentions NodeTool or asks for a workflow graph, a node, or a media generation run on their NodeTool server.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fplugins\u002Fchatgpt\u002Fnodetool\u002Fskills\u002Fnodetool\u002FSKILL.md",[90,93],{"categoryNo":91,"categoryName":92,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":94,"categoryName":95,"taxonomyType":96,"taxonomyTypeDesc":97},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[86,99,100,101,102,103,104,105,106,34,107,30,36,108,109,110,111,112,113,114,115,116],"MCP","Media","workflows","when","generation","media generation","with","visual","collections","work","debug","graph","validate","assets","run","server","mcp server","workflow",{"skillNo":118,"name":119,"nameZh":6,"description":120,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":121,"categories":122,"keywords":124,"publishedAt":46},"01M6BY4AEB05ZPBRVD078ATR2H","wan-2-6-prompting","Prompt Alibaba's Wan 2.6 video — text-to-video with a global style line plus timing-bracketed shots, and image-to-video that describes only what changes over time. Covers the shot-mode switch per provider, the prompt budget, negative prompts, motion intensity wording, audio input and prompt expansion. Use whenever the model id names a Wan 2.6 video route (alibaba\u002Fwan-2.6\u002Ftext-to-video, alibaba\u002Fwan-2.6\u002Fimage-to-video, wan-video\u002Fwan-2.6-t2v, wan-video\u002Fwan-2.6-i2v, wan-video\u002Fwan2.6-i2v-flash) on generate_video, animate_image or a TextToVideo node. Not for Wan 2.5 or 2.7, or the Wan 2.6 image routes.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fwan-2-6-prompting\u002FSKILL.md",[123],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[25,125,126,127,128,129,130,36,131,132,133,134,135,136,44,38,27,137,37,138,139],"wan 2 6 prompting","line","expansion","input","image-to-video","text-to-video","route","only","mode","time","flash","Budget","animate","Audio","Video",{"skillNo":141,"name":142,"nameZh":6,"description":143,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":144,"categories":145,"keywords":148,"publishedAt":46},"01M6BY47T4748ZG5ZMTAR97ZWZ","video-workflow","Build a reusable NodeTool video workflow for batches or repeated inputs. For a single finished piece, use a storyboard skill.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fvideo-workflow\u002FSKILL.md",[146,147],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":91,"categoryName":92,"taxonomyType":21,"taxonomyTypeDesc":22},[116,149,139,150,151,152,153,34,154,86,30],"video workflow","reusable","storyboard","build","batches","single",{"skillNo":156,"name":157,"nameZh":6,"description":158,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":159,"categories":160,"keywords":162,"publishedAt":46},"01M6BY4587N2SYN315SXGEMZC3","video-clone","Analyze a reference video and recreate its shot structure with a new product or cast in NodeTool.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fvideo-clone\u002FSKILL.md",[161],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[163,164,139,105,165,166,86,167,168,169,170],"clone","video clone","cast","new","analyze","product","structure","reference",{"skillNo":172,"name":173,"nameZh":6,"description":174,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":175,"categories":176,"keywords":178,"publishedAt":46},"01M6BY4257C2KQYYMD9EK6DK8E","video-audio-continuity","Keep sound continuous across a multi-scene piece cut from generated video — why one clip per scene hard-cuts the audio at every boundary, when to write all the scenes into a single generation instead, the sound-design vocabulary that survives a provider's copyright filter, and how to check the length, the aspect ratio and the mix of what came back. Use when a piece has more than one scene and the video model writes its own audio (Seedance 2, Veo 3, MiniMax H3, Kling 2.6 and later), or when a cut's sound drops out at the shot changes.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fvideo-audio-continuity\u002FSKILL.md",[177],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[179,139,138,180,181,182,183,184,185,186,187,188,189,102,190,32,37,191,192,193,40],"video audio continuity","continuity","write","why","back","veo 3","at","cut","later","across","filter","generated","mix","design","boundary"]