[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6PHADDV3QBVFBWS4W16DHWB":3,"skill-same:01M6PHADDV3QBVFBWS4W16DHWB":75},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":22,"publishedAt":43,"linkOnly":44,"markdownContent":45,"markdownBody":46,"frontMatter":47,"refName":48,"commitSha":49,"skillDirectory":50,"forksCount":51,"repositoryHtmlUrl":52,"repositoryHomepageUrl":6,"ownerHtmlUrl":53,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":54,"install":55,"resources":68},"01M6PHADDV3QBVFBWS4W16DHWB","anima-prompt-pool",null,"Build external prompt pools for Anima-based source generation. Use when creating many Anima prompts for later batch generation, especially when prompts must be saved in a stable JSON-friendly format with ids, categories, notes, subject tags, and natural-language scene text.","DevelopmentDummy\u002Fclaude-play","DevelopmentDummy","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F269318128?v=4","01M6PH7WG74ND790ED9698KRBP","01M6PH7WM7KC9N0V51ZCQ43P8B",3,"MIT","https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fdata\u002Ftools\u002Fcomfyui\u002Fskills\u002Fanima-prompt-pool\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42],"anima prompt pool","pool","prompt","anima","generation","for","natural-language","saved","categories","with","tags","notes","build","later","use","text","format","json","ids","batch","2026-11-10 10:12:30",false,"---\nname: anima-prompt-pool\ndescription: Build external prompt pools for Anima-based source generation. Use when creating many Anima prompts for later batch generation, especially when prompts must be saved in a stable JSON-friendly format with ids, categories, notes, subject tags, and natural-language scene text.\n---\n\n# Anima Prompt Pool\n\n## Overview\n\nCreate prompt-bank entries that are easy to append into the batch pipeline later.\nWrite prompts for Anima source generation, not for the Illustrious teacher pass.\n\n## Core Rules\n\n- Use `subject_tags + scene_prompt` as the canonical format.\n- `subject_tags` holds visible anchors in tag form.\n- `scene_prompt` holds the main natural-language scene description.\n- Optimize for batch consistency, not literary flair.\n- Keep one entry focused on one scene idea.\n- If an outfit, lighting, or camera angle matters, state it explicitly.\n\n## Output Format\n\nDefault to this JSON object format:\n\n```json\n{\n  \"id\": \"p001\",\n  \"category\": [\"office\", \"professional\", \"glasses\"],\n  \"exposure\": \"clothed\",\n  \"subject_tags\": \"1girl, solo, dark navy hair, long hair, teal eyes, pale skin, thin silver-rim glasses, white research blouse with rolled-up sleeves, dark navy vest, black pencil skirt, black tights, sitting at desk, legs crossed, soft skin, cool blue tones, monitor glow, upper body\",\n  \"scene_prompt\": \"An adult woman with long dark navy hair and teal eyes is sitting at a sleek modern office desk, wearing a white research blouse with rolled-up sleeves over a dark navy vest and black pencil skirt with black tights. She has thin silver-rimmed glasses and is focusing on a holographic display. The scene is lit by cool blue screen glow and soft ambient office lights. The image should feel professional and focused, with sharp suit textures and polished glass surfaces.\",\n  \"negative_prompt\": \"multiple girls, 2girls, extra person, crowd, people in background\",\n  \"notes\": \"Optional short note about intent or risk.\"\n}\n```\n\n### exposure field\n\n`exposure` classifies the scene's nudity level. This drives which body\u002Fexposure tags MUST appear in `subject_tags` and ensures caption–image consistency for LoRA training.\n\n| exposure | Definition | Required tags (subject_tags) | Forbidden tags (subject_tags) |\n|----------|-----------|------------------------------|-------------------------------|\n| `clothed` | Fully dressed, no sexual parts visible | Explicit clothing items (`white blouse`, `black skirt`, etc.), `clothed` | `nude`, `nipples`, `pussy`, `bare_breasts`, `topless` |\n| `partial` | Partially exposed — underwear visible, cleavage, sideboob, lifted skirt, see-through, etc. | Specific exposure state (`cleavage`, `black lace bra`, `unbuttoned shirt`, `sideboob`, `lifted skirt`, `panties visible`, etc.) | `nude`, `fully_clothed` |\n| `nude` | Fully nude or near-nude, sexual parts directly visible | `nude` or `topless`, plus visible anatomy (`nipples`, `navel`, `bare_breasts`, `pussy`, etc. as applicable to the composition) | `clothed`, `covered_nipples`, `fully_clothed` |\n\n**Rules:**\n- Every prompt MUST have an `exposure` field. No default — force explicit classification.\n- The tags in `subject_tags` MUST be consistent with the declared `exposure`. If `exposure` is `clothed`, do not include `nipples`. If `exposure` is `nude`, do not include `clothed`.\n- For `partial`, specify exactly what is exposed and what remains covered. Ambiguity here causes the worst training noise.\n- `exposure` value is also carried into `category` automatically for filtering (e.g., `[\"bedroom\", \"nude\"]`).\n\n### negative_prompt field\n\n- `negative_prompt` is optional per-item. If omitted, the pipeline uses the global default from `pipeline_meta.json` (`source_default_negative` \u002F `teacher_default_negative`).\n- Use per-item negatives only when a scene needs **extra** suppression beyond the global default (e.g., a crowded street scene that tends to spawn extra characters, or a mirror scene that generates duplicates).\n- Common per-item negatives:\n  - Crowded environments: `multiple girls, 2girls, extra person, crowd, people in background`\n  - Mirror\u002Freflection scenes: `duplicate, clone, symmetry, reflection`\n  - Indoor scenes with windows: `extra person, crowd, people in background`\n- Do **not** duplicate quality negatives (`worst quality, low quality`) here — those are already in the global default.\n- For multi-character scenes (2girls etc.), **remove** the anti-multi tags from negative: `\"negative_prompt\": \"\"` to override the global default.\n\nWhen asked for many prompts, output a JSON array of those objects.\n\nLegacy compatibility:\n\n- old `prompt` single-field format may still exist\n- but new pools should default to `subject_tags + scene_prompt`\n\n## Real File Example\n\nUse a `prompt-bank.json` file like this:\n\n```json\n[\n  {\n    \"id\": \"p001\",\n    \"category\": [\"office\", \"professional\", \"glasses\", \"clothed\"],\n    \"exposure\": \"clothed\",\n    \"subject_tags\": \"1girl, solo, dark navy hair, long hair, teal eyes, pale skin, thin silver-rim glasses, white research blouse with rolled-up sleeves, dark navy vest, black pencil skirt, black tights, sitting at desk, legs crossed, clothed, soft skin, cool blue tones, monitor glow, upper body\",\n    \"scene_prompt\": \"An adult woman with long dark navy hair and teal eyes is sitting at a sleek modern office desk, wearing a white research blouse with rolled-up sleeves layered under a dark navy vest, paired with a black pencil skirt and black tights. She wears thin silver-rimmed glasses. She is focusing on a holographic display. The scene is lit by cool blue screen glow and soft ambient office lights. The image should feel professional and focused, with sharp suit textures and polished glass surfaces.\",\n    \"notes\": \"Good for polished interior surfaces.\"\n  },\n  {\n    \"id\": \"p002\",\n    \"category\": [\"nature\", \"forest\", \"ethereal\", \"clothed\"],\n    \"exposure\": \"clothed\",\n    \"subject_tags\": \"1girl, solo, dark navy hair, long hair, teal eyes, pale skin, white cotton sundress, thin shoulder straps, barefoot, walking, clothed, soft skin, warm tones, dappled sunlight, full body\",\n    \"scene_prompt\": \"An adult woman with long dark navy hair and teal eyes is walking through a sun-drenched forest, wearing a white cotton sundress with thin shoulder straps. She is barefoot on the mossy ground. The scene is lit by warm dapples of sunlight filtering through the canopy of ancient mossy trees with floating pollen. The image should feel ethereal and serene, with soft fabric movement and detailed leaf textures.\",\n    \"notes\": \"Good for soft natural light.\"\n  },\n  {\n    \"id\": \"p003\",\n    \"category\": [\"bedroom\", \"intimate\", \"partial\"],\n    \"exposure\": \"partial\",\n    \"subject_tags\": \"1girl, solo, dark navy hair, long hair, teal eyes, pale skin, unbuttoned white dress shirt, black lace bra visible, cleavage, bare legs, sitting on bed edge, leaning back on hands, soft skin, warm tones, dim lamp light, full body\",\n    \"scene_prompt\": \"An adult woman with long dark navy hair and teal eyes is sitting on the edge of a bed, wearing an unbuttoned white dress shirt that reveals a black lace bra underneath. Her legs are bare. She is leaning back on her hands with a relaxed expression. The scene is lit by warm dim lamp light from a bedside table. The image should feel intimate and natural, with soft fabric folds and warm skin tones.\",\n    \"notes\": \"Partial exposure — bra visible through unbuttoned shirt.\"\n  },\n  {\n    \"id\": \"p004\",\n    \"category\": [\"bathroom\", \"morning\", \"nude\"],\n    \"exposure\": \"nude\",\n    \"subject_tags\": \"1girl, solo, dark navy hair, long hair, wet hair, teal eyes, pale skin, nude, bare breasts, nipples, navel, standing, holding towel at side, wet skin, water droplets, soft skin, warm tones, soft morning light, full body\",\n    \"scene_prompt\": \"An adult woman with long dark navy hair (wet and clinging to her shoulders) and teal eyes is standing nude in a steamy bathroom, holding a white towel casually at her side. Water droplets glisten on her pale skin. The scene is lit by soft warm morning light coming through a frosted window. The image should feel natural and candid, with realistic water droplet textures and steam diffusion.\",\n    \"notes\": \"Full nude — all relevant anatomy tags included.\"\n  }\n]\n```\n\nUse a `review-queue.json` file like this:\n\n```json\n[\n  {\n    \"id\": \"p001\",\n    \"status\": \"pending\",\n    \"source_image\": \"images\u002Fsource\u002Fp001.png\",\n    \"teacher_image\": \"images\u002Fteacher\u002Fp001.png\",\n    \"subject_tags\": \"1girl, solo, long navy hair, silver-rimmed glasses, white research blouse, dark navy vest, black pleated skirt\",\n    \"scene_prompt\": \"An adult woman is sitting at a sleek modern office desk, focusing on a holographic display. The scene is lit by cool blue screen glow and soft ambient office lights. The image should feel professional and focused, with sharp suit textures and polished glass surfaces.\",\n    \"category\": [\"office\", \"professional\", \"glasses\"],\n    \"notes\": \"Good for polished interior surfaces.\"\n  }\n]\n```\n\nRecommended status values:\n\n- `pending` — not reviewed yet\n- `yes` — approved for dataset\n- `no` — rejected\n- `hold` — keep for later review\n\n## Prompt Construction Workflow\n\nFor each prompt:\n\n1. Choose one clear scene family\n   - neon city\n   - luxury interior\n   - futuristic stage\n   - bedroom\n   - bathroom\n   - office\n   - fantasy tavern\n\n2. Build `subject_tags` with all six axes filled\n   subject_tags is used BOTH for Anima source generation AND as the Illustrious teacher prompt (without natural language). It must be rich enough for the teacher to render the scene correctly on its own.\n\n   **Required axes (always include all six):**\n\n   | Axis | Examples | Why |\n   |------|---------|-----|\n   | **Character identity** | `1girl, solo, dark navy hair, long hair, teal eyes, pale skin` | Teacher doesn't know the character otherwise |\n   | **Outfit \u002F exposure** | `white sundress, barefoot, clothed` or `nude, bare breasts, nipples` or `unbuttoned shirt, black lace bra visible, cleavage` — **must match the `exposure` field** | Ambiguous nudity level causes rendering errors AND training noise |\n   | **Pose \u002F anatomy** | `standing, arms at sides` or `on back, lying on bed, knees up, navel` | Prevents body distortion, especially in lying\u002Fcrouching poses |\n   | **Skin \u002F texture state** | `soft skin` or `sweating, wet skin` or `oiled skin, sweat droplets` | Directly affects erotic scene quality |\n   | **Color \u002F lighting tone** | `warm tones, golden light` or `cool blue tones, dim lighting` or `high contrast, neon glow` | Anchors the teacher's color grading direction |\n   | **Composition** | `upper body` or `full body` or `close-up, face focus` | Prevents framing mismatch between source and teacher |\n\n   **Example — bedroom nude (exposure: \"nude\"):**\n   ```\n   1girl, solo, dark navy hair, long hair, teal eyes, pale skin, nude, bare breasts, nipples, navel, flat stomach, on back, lying on bed, arms above head, soft skin, messy hair, warm tones, dim lighting, full body\n   ```\n\n   **Example — office clothed (exposure: \"clothed\"):**\n   ```\n   1girl, solo, dark navy hair, long hair, teal eyes, pale skin, thin silver-rim glasses, white research blouse with rolled-up sleeves, dark navy vest, black pencil skirt, black tights, clothed, sitting at desk, legs crossed, soft skin, cool blue tones, monitor glow, upper body\n   ```\n\n   **Example — bedroom partial (exposure: \"partial\"):**\n   ```\n   1girl, solo, dark navy hair, long hair, teal eyes, pale skin, oversized white t-shirt, no bra, sideboob, black panties, bare legs, sitting on bed, knees up, soft skin, warm tones, morning light, full body\n   ```\n\n3. Write `scene_prompt` as natural language with explicit visual details\n   - `scene_prompt`: action, environment, lighting, mood, material richness\n   - This is used by Anima only (not teacher), so it can be descriptive prose\n   - **MUST explicitly describe**: hair color & style, clothing items with color\u002Fmaterial\u002Fdetail, accessories, footwear\n   - The first sentence should establish the character's appearance: *\"An adult woman with long dark navy hair and teal eyes is...\"*\n   - Clothing descriptions should be specific: not just \"wearing a dress\" but \"wearing a white cotton sundress with thin shoulder straps\"\n   - If exposure is `partial` or `nude`, describe the state of dress\u002Fundress naturally in the prose\n   - Accessories (glasses, jewelry, hair ties, etc.) should be mentioned if present\n\n4. Decide if per-item negative is needed\n   - Most scenes: omit `negative_prompt` (global default handles it)\n   - Busy backgrounds (streets, crowds, festivals): add `\"negative_prompt\": \"multiple girls, 2girls, extra person, crowd, people in background\"`\n   - Multi-character scenes: add `\"negative_prompt\": \"\"` to clear the global anti-multi default\n   - Mirror\u002Fwindow scenes: add `\"negative_prompt\": \"duplicate, clone, reflection\"`\n\n5. Add a short note only if needed\n   - good for glossy clothing\n   - strong color contrast\n   - risk of cluttered background\n\n## Writing Pattern\n\nUse this skeleton:\n\n```text\nexposure: [clothed | partial | nude]\nsubject_tags: [인원], [캐릭터 외형: 머리색, 머리길이, 눈색, 피부], [의상 또는 노출 상태 — exposure와 일치], [포즈\u002F해부학], [피부 질감], [컬러\u002F라이팅 톤], [구도]\nscene_prompt: An adult woman with [hair color & style] and [eye color] is [pose\u002Faction] in [environment], wearing [specific clothing with color\u002Fmaterial\u002Fdetail OR nude state description]. [Accessories if any]. The scene is lit by [lighting]. The image should feel [mood], with [material\u002Fcolor\u002Fcomposition details].\nnegative_prompt: (optional, only if scene background is busy or multi-character suppression is needed)\n```\n\n**Checklist before submitting a prompt — all seven items present?**\n- [ ] `exposure` field set (clothed \u002F partial \u002F nude)\n- [ ] Character identity (hair color & style, eyes, skin)\n- [ ] Outfit or exposure tags consistent with `exposure` field\n- [ ] Pose and anatomy anchors\n- [ ] Skin\u002Ftexture state\n- [ ] Color\u002Flighting tone\n- [ ] Composition (framing)\n- [ ] scene_prompt explicitly describes hair, clothing\u002Fnudity, and accessories in natural language\n\n## Constraints\n\n- Prefer prompts that remain readable after batch generation.\n- Do not overload every prompt with too many visual ideas.\n- Do not write teacher-stage trigger tags here unless explicitly requested.\n- Do not convert the whole entry into tags only; keep `scene_prompt` in natural language.\n- If the user wants erotic prompts, keep the same structured format and focus on scene clarity.\n\n## If the User Requests Variations\n\nGenerate variations by changing only one or two axes:\n\n- environment\n- lighting\n- outfit material\n- color palette\n- camera distance\n\nDo not change every axis at once unless the user asks for maximum variety.\n\n## Deliverables\n\nWhen asked to prepare a pool, provide:\n\n1. A JSON array\n2. Optional category summary\n3. Optional warnings about prompts likely to fail in batch\n","# Anima Prompt Pool\n\n## Overview\n\nCreate prompt-bank entries that are easy to append into the batch pipeline later.\nWrite prompts for Anima source generation, not for the Illustrious teacher pass.\n\n## Core Rules\n\n- Use `subject_tags + scene_prompt` as the canonical format.\n- `subject_tags` holds visible anchors in tag form.\n- `scene_prompt` holds the main natural-language scene description.\n- Optimize for batch consistency, not literary flair.\n- Keep one entry focused on one scene idea.\n- If an outfit, lighting, or camera angle matters, state it explicitly.\n\n## Output Format\n\nDefault to this JSON object format:\n\n```json\n{\n  \"id\": \"p001\",\n  \"category\": [\"office\", \"professional\", \"glasses\"],\n  \"exposure\": \"clothed\",\n  \"subject_tags\": \"1girl, solo, dark navy hair, long hair, teal eyes, pale skin, thin silver-rim glasses, white research blouse with rolled-up sleeves, dark navy vest, black pencil skirt, black tights, sitting at desk, legs crossed, soft skin, cool blue tones, monitor glow, upper body\",\n  \"scene_prompt\": \"An adult woman with long dark navy hair and teal eyes is sitting at a sleek modern office desk, wearing a white research blouse with rolled-up sleeves over a dark navy vest and black pencil skirt with black tights. She has thin silver-rimmed glasses and is focusing on a holographic display. The scene is lit by cool blue screen glow and soft ambient office lights. The image should feel professional and focused, with sharp suit textures and polished glass surfaces.\",\n  \"negative_prompt\": \"multiple girls, 2girls, extra person, crowd, people in background\",\n  \"notes\": \"Optional short note about intent or risk.\"\n}\n```\n\n### exposure field\n\n`exposure` classifies the scene's nudity level. This drives which body\u002Fexposure tags MUST appear in `subject_tags` and ensures caption–image consistency for LoRA training.\n\n| exposure | Definition | Required tags (subject_tags) | Forbidden tags (subject_tags) |\n|----------|-----------|------------------------------|-------------------------------|\n| `clothed` | Fully dressed, no sexual parts visible | Explicit clothing items (`white blouse`, `black skirt`, etc.), `clothed` | `nude`, `nipples`, `pussy`, `bare_breasts`, `topless` |\n| `partial` | Partially exposed — underwear visible, cleavage, sideboob, lifted skirt, see-through, etc. | Specific exposure state (`cleavage`, `black lace bra`, `unbuttoned shirt`, `sideboob`, `lifted skirt`, `panties visible`, etc.) | `nude`, `fully_clothed` |\n| `nude` | Fully nude or near-nude, sexual parts directly visible | `nude` or `topless`, plus visible anatomy (`nipples`, `navel`, `bare_breasts`, `pussy`, etc. as applicable to the composition) | `clothed`, `covered_nipples`, `fully_clothed` |\n\n**Rules:**\n- Every prompt MUST have an `exposure` field. No default — force explicit classification.\n- The tags in `subject_tags` MUST be consistent with the declared `exposure`. If `exposure` is `clothed`, do not include `nipples`. If `exposure` is `nude`, do not include `clothed`.\n- For `partial`, specify exactly what is exposed and what remains covered. Ambiguity here causes the worst training noise.\n- `exposure` value is also carried into `category` automatically for filtering (e.g., `[\"bedroom\", \"nude\"]`).\n\n### negative_prompt field\n\n- `negative_prompt` is optional per-item. If omitted, the pipeline uses the global default from `pipeline_meta.json` (`source_default_negative` \u002F `teacher_default_negative`).\n- Use per-item negatives only when a scene needs **extra** suppression beyond the global default (e.g., a crowded street scene that tends to spawn extra characters, or a mirror scene that generates duplicates).\n- Common per-item negatives:\n  - Crowded environments: `multiple girls, 2girls, extra person, crowd, people in background`\n  - Mirror\u002Freflection scenes: `duplicate, clone, symmetry, reflection`\n  - Indoor scenes with windows: `extra person, crowd, people in background`\n- Do **not** duplicate quality negatives (`worst quality, low quality`) here — those are already in the global default.\n- For multi-character scenes (2girls etc.), **remove** the anti-multi tags from negative: `\"negative_prompt\": \"\"` to override the global default.\n\nWhen asked for many prompts, output a JSON array of those objects.\n\nLegacy compatibility:\n\n- old `prompt` single-field format may still exist\n- but new pools should default to `subject_tags + scene_prompt`\n\n## Real File Example\n\nUse a `prompt-bank.json` file like this:\n\n```json\n[\n  {\n    \"id\": \"p001\",\n    \"category\": [\"office\", \"professional\", \"glasses\", \"clothed\"],\n    \"exposure\": \"clothed\",\n    \"subject_tags\": \"1girl, solo, dark navy hair, long hair, teal eyes, pale skin, thin silver-rim glasses, white research blouse with rolled-up sleeves, dark navy vest, black pencil skirt, black tights, sitting at desk, legs crossed, clothed, soft skin, cool blue tones, monitor glow, upper body\",\n    \"scene_prompt\": \"An adult woman with long dark navy hair and teal eyes is sitting at a sleek modern office desk, wearing a white research blouse with rolled-up sleeves layered under a dark navy vest, paired with a black pencil skirt and black tights. She wears thin silver-rimmed glasses. She is focusing on a holographic display. The scene is lit by cool blue screen glow and soft ambient office lights. The image should feel professional and focused, with sharp suit textures and polished glass surfaces.\",\n    \"notes\": \"Good for polished interior surfaces.\"\n  },\n  {\n    \"id\": \"p002\",\n    \"category\": [\"nature\", \"forest\", \"ethereal\", \"clothed\"],\n    \"exposure\": \"clothed\",\n    \"subject_tags\": \"1girl, solo, dark navy hair, long hair, teal eyes, pale skin, white cotton sundress, thin shoulder straps, barefoot, walking, clothed, soft skin, warm tones, dappled sunlight, full body\",\n    \"scene_prompt\": \"An adult woman with long dark navy hair and teal eyes is walking through a sun-drenched forest, wearing a white cotton sundress with thin shoulder straps. She is barefoot on the mossy ground. The scene is lit by warm dapples of sunlight filtering through the canopy of ancient mossy trees with floating pollen. The image should feel ethereal and serene, with soft fabric movement and detailed leaf textures.\",\n    \"notes\": \"Good for soft natural light.\"\n  },\n  {\n    \"id\": \"p003\",\n    \"category\": [\"bedroom\", \"intimate\", \"partial\"],\n    \"exposure\": \"partial\",\n    \"subject_tags\": \"1girl, solo, dark navy hair, long hair, teal eyes, pale skin, unbuttoned white dress shirt, black lace bra visible, cleavage, bare legs, sitting on bed edge, leaning back on hands, soft skin, warm tones, dim lamp light, full body\",\n    \"scene_prompt\": \"An adult woman with long dark navy hair and teal eyes is sitting on the edge of a bed, wearing an unbuttoned white dress shirt that reveals a black lace bra underneath. Her legs are bare. She is leaning back on her hands with a relaxed expression. The scene is lit by warm dim lamp light from a bedside table. The image should feel intimate and natural, with soft fabric folds and warm skin tones.\",\n    \"notes\": \"Partial exposure — bra visible through unbuttoned shirt.\"\n  },\n  {\n    \"id\": \"p004\",\n    \"category\": [\"bathroom\", \"morning\", \"nude\"],\n    \"exposure\": \"nude\",\n    \"subject_tags\": \"1girl, solo, dark navy hair, long hair, wet hair, teal eyes, pale skin, nude, bare breasts, nipples, navel, standing, holding towel at side, wet skin, water droplets, soft skin, warm tones, soft morning light, full body\",\n    \"scene_prompt\": \"An adult woman with long dark navy hair (wet and clinging to her shoulders) and teal eyes is standing nude in a steamy bathroom, holding a white towel casually at her side. Water droplets glisten on her pale skin. The scene is lit by soft warm morning light coming through a frosted window. The image should feel natural and candid, with realistic water droplet textures and steam diffusion.\",\n    \"notes\": \"Full nude — all relevant anatomy tags included.\"\n  }\n]\n```\n\nUse a `review-queue.json` file like this:\n\n```json\n[\n  {\n    \"id\": \"p001\",\n    \"status\": \"pending\",\n    \"source_image\": \"images\u002Fsource\u002Fp001.png\",\n    \"teacher_image\": \"images\u002Fteacher\u002Fp001.png\",\n    \"subject_tags\": \"1girl, solo, long navy hair, silver-rimmed glasses, white research blouse, dark navy vest, black pleated skirt\",\n    \"scene_prompt\": \"An adult woman is sitting at a sleek modern office desk, focusing on a holographic display. The scene is lit by cool blue screen glow and soft ambient office lights. The image should feel professional and focused, with sharp suit textures and polished glass surfaces.\",\n    \"category\": [\"office\", \"professional\", \"glasses\"],\n    \"notes\": \"Good for polished interior surfaces.\"\n  }\n]\n```\n\nRecommended status values:\n\n- `pending` — not reviewed yet\n- `yes` — approved for dataset\n- `no` — rejected\n- `hold` — keep for later review\n\n## Prompt Construction Workflow\n\nFor each prompt:\n\n1. Choose one clear scene family\n   - neon city\n   - luxury interior\n   - futuristic stage\n   - bedroom\n   - bathroom\n   - office\n   - fantasy tavern\n\n2. Build `subject_tags` with all six axes filled\n   subject_tags is used BOTH for Anima source generation AND as the Illustrious teacher prompt (without natural language). It must be rich enough for the teacher to render the scene correctly on its own.\n\n   **Required axes (always include all six):**\n\n   | Axis | Examples | Why |\n   |------|---------|-----|\n   | **Character identity** | `1girl, solo, dark navy hair, long hair, teal eyes, pale skin` | Teacher doesn't know the character otherwise |\n   | **Outfit \u002F exposure** | `white sundress, barefoot, clothed` or `nude, bare breasts, nipples` or `unbuttoned shirt, black lace bra visible, cleavage` — **must match the `exposure` field** | Ambiguous nudity level causes rendering errors AND training noise |\n   | **Pose \u002F anatomy** | `standing, arms at sides` or `on back, lying on bed, knees up, navel` | Prevents body distortion, especially in lying\u002Fcrouching poses |\n   | **Skin \u002F texture state** | `soft skin` or `sweating, wet skin` or `oiled skin, sweat droplets` | Directly affects erotic scene quality |\n   | **Color \u002F lighting tone** | `warm tones, golden light` or `cool blue tones, dim lighting` or `high contrast, neon glow` | Anchors the teacher's color grading direction |\n   | **Composition** | `upper body` or `full body` or `close-up, face focus` | Prevents framing mismatch between source and teacher |\n\n   **Example — bedroom nude (exposure: \"nude\"):**\n   ```\n   1girl, solo, dark navy hair, long hair, teal eyes, pale skin, nude, bare breasts, nipples, navel, flat stomach, on back, lying on bed, arms above head, soft skin, messy hair, warm tones, dim lighting, full body\n   ```\n\n   **Example — office clothed (exposure: \"clothed\"):**\n   ```\n   1girl, solo, dark navy hair, long hair, teal eyes, pale skin, thin silver-rim glasses, white research blouse with rolled-up sleeves, dark navy vest, black pencil skirt, black tights, clothed, sitting at desk, legs crossed, soft skin, cool blue tones, monitor glow, upper body\n   ```\n\n   **Example — bedroom partial (exposure: \"partial\"):**\n   ```\n   1girl, solo, dark navy hair, long hair, teal eyes, pale skin, oversized white t-shirt, no bra, sideboob, black panties, bare legs, sitting on bed, knees up, soft skin, warm tones, morning light, full body\n   ```\n\n3. Write `scene_prompt` as natural language with explicit visual details\n   - `scene_prompt`: action, environment, lighting, mood, material richness\n   - This is used by Anima only (not teacher), so it can be descriptive prose\n   - **MUST explicitly describe**: hair color & style, clothing items with color\u002Fmaterial\u002Fdetail, accessories, footwear\n   - The first sentence should establish the character's appearance: *\"An adult woman with long dark navy hair and teal eyes is...\"*\n   - Clothing descriptions should be specific: not just \"wearing a dress\" but \"wearing a white cotton sundress with thin shoulder straps\"\n   - If exposure is `partial` or `nude`, describe the state of dress\u002Fundress naturally in the prose\n   - Accessories (glasses, jewelry, hair ties, etc.) should be mentioned if present\n\n4. Decide if per-item negative is needed\n   - Most scenes: omit `negative_prompt` (global default handles it)\n   - Busy backgrounds (streets, crowds, festivals): add `\"negative_prompt\": \"multiple girls, 2girls, extra person, crowd, people in background\"`\n   - Multi-character scenes: add `\"negative_prompt\": \"\"` to clear the global anti-multi default\n   - Mirror\u002Fwindow scenes: add `\"negative_prompt\": \"duplicate, clone, reflection\"`\n\n5. Add a short note only if needed\n   - good for glossy clothing\n   - strong color contrast\n   - risk of cluttered background\n\n## Writing Pattern\n\nUse this skeleton:\n\n```text\nexposure: [clothed | partial | nude]\nsubject_tags: [인원], [캐릭터 외형: 머리색, 머리길이, 눈색, 피부], [의상 또는 노출 상태 — exposure와 일치], [포즈\u002F해부학], [피부 질감], [컬러\u002F라이팅 톤], [구도]\nscene_prompt: An adult woman with [hair color & style] and [eye color] is [pose\u002Faction] in [environment], wearing [specific clothing with color\u002Fmaterial\u002Fdetail OR nude state description]. [Accessories if any]. The scene is lit by [lighting]. The image should feel [mood], with [material\u002Fcolor\u002Fcomposition details].\nnegative_prompt: (optional, only if scene background is busy or multi-character suppression is needed)\n```\n\n**Checklist before submitting a prompt — all seven items present?**\n- [ ] `exposure` field set (clothed \u002F partial \u002F nude)\n- [ ] Character identity (hair color & style, eyes, skin)\n- [ ] Outfit or exposure tags consistent with `exposure` field\n- [ ] Pose and anatomy anchors\n- [ ] Skin\u002Ftexture state\n- [ ] Color\u002Flighting tone\n- [ ] Composition (framing)\n- [ ] scene_prompt explicitly describes hair, clothing\u002Fnudity, and accessories in natural language\n\n## Constraints\n\n- Prefer prompts that remain readable after batch generation.\n- Do not overload every prompt with too many visual ideas.\n- Do not write teacher-stage trigger tags here unless explicitly requested.\n- Do not convert the whole entry into tags only; keep `scene_prompt` in natural language.\n- If the user wants erotic prompts, keep the same structured format and focus on scene clarity.\n\n## If the User Requests Variations\n\nGenerate variations by changing only one or two axes:\n\n- environment\n- lighting\n- outfit material\n- color palette\n- camera distance\n\nDo not change every axis at once unless the user asks for maximum variety.\n\n## Deliverables\n\nWhen asked to prepare a pool, provide:\n\n1. A JSON array\n2. Optional category summary\n3. Optional warnings about prompts likely to fail in batch\n",{"name":5,"description":7},"main","87c4e526fb2dfb2c6bfeeb9f901d6320fcb457ec","data\u002Ftools\u002Fcomfyui\u002Fskills\u002Fanima-prompt-pool",0,"https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play","https:\u002F\u002Fgithub.com\u002FDevelopmentDummy","2026-10-07 22:33:19",{"source":8,"skillName":5,"commands":56,"prompt":66,"zipAvailable":67},[57,60,63],{"tool":58,"command":59},"npx","npx skills add DevelopmentDummy\u002Fclaude-play --skill anima-prompt-pool",{"tool":61,"command":62},"bunx","bunx skills add DevelopmentDummy\u002Fclaude-play --skill anima-prompt-pool",{"tool":64,"command":65},"pnpm","pnpm dlx skills add DevelopmentDummy\u002Fclaude-play --skill anima-prompt-pool","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fdata\u002Ftools\u002Fcomfyui\u002Fskills\u002Fanima-prompt-pool\u002FSKILL.md",true,[69],{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"agents\u002Fopenai.yaml","REFERENCE","参考文档",191,"https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fdata\u002Ftools\u002Fcomfyui\u002Fskills\u002Fanima-prompt-pool\u002Fagents\u002Fopenai.yaml",[76,97,111,130,140,149],{"skillNo":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":86,"publishedAt":43},"01M6PHD8Y8CNRBG3FR2ZAHWX5J","generate-video","claude-play-bridge MCP 서버로 영상을 생성한다. 로컬 ComfyUI의 영상 워크플로 패키지(MiniMax H3 = 영상+오디오 동시 생성, WAN 2.2 i2v, Z-Image→영상)를 호출한다. 렌더가 수 분~수십 분 걸리므로 async 큐잉 + 파일 폴링이 기본 절차다. \"영상\", \"동영상\", \"비디오\", \"video\", \"i2v\", \"t2v\", \"클립 만들어\", \"움직이는\", \"애니메이션 영상\" 요청에 트리거된다.","https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fscripts\u002Fexternal-package\u002Fskills\u002Fgenerate-video\u002FSKILL.md",[82,85],{"categoryNo":83,"categoryName":84,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000016","图像与多媒体",{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[87,88,89,90,91,92,93,94,95,96],"Video","generate","generate video","MCP","image","play","bridge","minimax","async","minimax h3",{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":105,"publishedAt":43},"01M6PHD3MAQ9QYQQFG50WBP5JZ","generate-image","claude-play-bridge MCP 서버로 이미지를 생성한다. 기본 백엔드는 ComfyUI(로컬 GPU, 워크플로 패키지 기반)이고, 사용자가 명시적으로 요청할 때만 GPT\u002FGemini 백엔드를 사용한다.","https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fscripts\u002Fexternal-package\u002Fskills\u002Fgenerate-image\u002FSKILL.md",[103,104],{"categoryNo":83,"categoryName":84,"taxonomyType":20,"taxonomyTypeDesc":21},{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[106,88,91,93,90,107,92,108,109,110],"generate image","gemini","gpu","gpt","comfyui",{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":118,"publishedAt":43},"01M6PHD0134XK1NQNK4YMZB2E1","generate-image-openai","OpenAI\u002FGPT 이미지 생성(GPT Image 품질 \u002F 텍스트 렌더링·이미지 편집 강점). 기본 백엔드는 Codex CLI 구독(건당 과금 없음). ⚠️ 사용자가 이 백엔드를 명시적으로 지정했을 때만 사용한다(예 \"GPT로\u002FOpenAI로 그려줘\"). 기본 이미지 생성은 generate-image(ComfyUI)이며, 사용자가 콕 집어 요청하지 않았다면 절대 OpenAI를 자동으로 선택하지 마라.","https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fdata\u002Ftools\u002Fopenai\u002Fskills\u002Fgenerate-image-openai\u002FSKILL.md",[117],{"categoryNo":83,"categoryName":84,"taxonomyType":20,"taxonomyTypeDesc":21},[119,106,120,88,91,121,122,123,124,125,109,126,127,128,110,129],"generate image openai","openai","렌더링","cli","이미지","codex","codex cli","Bash","read","write","텍스트",{"skillNo":131,"name":132,"nameZh":6,"description":133,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":134,"categories":135,"keywords":136,"publishedAt":43},"01M6PHCTMBPTG56KMK8VH5SCTE","generate-spritesheet","Gemini 이미지 생성으로 게임용 아이콘 스프라이트시트를 만들고 자동 슬라이스한다. 아이콘, UI 리소스, 뱃지, 마커 등 여러 개의 동일 카테고리 에셋이 필요할 때 사용한다.","https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fdata\u002Ftools\u002Fgemini\u002Fskills\u002Fgenerate-spritesheet\u002FSKILL.md",[],[137,88,107,138,126,127,128,139,123],"generate spritesheet","ui","edit",{"skillNo":141,"name":142,"nameZh":6,"description":143,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":144,"categories":145,"keywords":147,"publishedAt":43},"01M6PHCPR07KYS2RPNZER9WS2S","generate-image-gemini","Gemini API로 이미지를 생성한다. 패널 배경, UI 리소스, 아이콘, 장면 삽화 등 디자인 리소스 제작에 사용한다. ComfyUI가 없을 때의 이미지 생성 대안이기도 하다.","https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fdata\u002Ftools\u002Fgemini\u002Fskills\u002Fgenerate-image-gemini\u002FSKILL.md",[146],{"categoryNo":83,"categoryName":84,"taxonomyType":20,"taxonomyTypeDesc":21},[148,107,91,88,106,128,127,126,138,123],"generate image gemini",{"skillNo":150,"name":151,"nameZh":6,"description":152,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":153,"categories":154,"keywords":159,"publishedAt":43},"01M6PHCJD7S7KT64BB9F215TP6","workflow-research","새 ComfyUI 워크플로우를 설계하거나 신규 모델, 양자화 모델, LoRA 개발 전략을 비교할 때 분석 틀을 잡는다.","https:\u002F\u002Fgithub.com\u002FDevelopmentDummy\u002Fclaude-play\u002Fblob\u002Fmain\u002Fdata\u002Ftools\u002Fcomfyui\u002Fskills\u002Fworkflow-research\u002FSKILL.md",[155,158],{"categoryNo":156,"categoryName":157,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000021","资料检索与研究",{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[160,161,162,110,127,163],"workflow","research","workflow research","LoRA"]