{
  "schemaVersion": "1.0",
  "item": {
    "slug": "anima",
    "name": "Anima",
    "source": "tencent",
    "type": "skill",
    "category": "AI 智能",
    "sourceUrl": "https://clawhub.ai/HMyaoyuan/anima",
    "canonicalUrl": "https://clawhub.ai/HMyaoyuan/anima",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/anima",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=anima",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "ASSETS_PLAN.md",
      "SKILL.md",
      "assets/manifest.json",
      "assets/production_plan.csv",
      "package.json",
      "run.js"
    ],
    "primaryDoc": "SKILL.md",
    "quickSetup": [
      "Download the package from Yavira.",
      "Extract the archive and review SKILL.md first.",
      "Import or place the package into your OpenClaw setup."
    ],
    "agentAssist": {
      "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
      "steps": [
        "Download the package from Yavira.",
        "Extract it into a folder your agent can access.",
        "Paste one of the prompts below and point your agent at the extracted folder."
      ],
      "prompts": [
        {
          "label": "New install",
          "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
        },
        {
          "label": "Upgrade existing",
          "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
        }
      ]
    },
    "sourceHealth": {
      "source": "tencent",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-04-23T16:43:11.935Z",
      "expiresAt": "2026-04-30T16:43:11.935Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard",
        "contentDisposition": "attachment; filename=\"4claw-imageboard-1.0.1.zip\"",
        "redirectLocation": null,
        "bodySnippet": null
      },
      "scope": "source",
      "summary": "Source download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this source.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/anima"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    },
    "downloadPageUrl": "https://openagent3.xyz/downloads/anima",
    "agentPageUrl": "https://openagent3.xyz/skills/anima/agent",
    "manifestUrl": "https://openagent3.xyz/skills/anima/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/anima/agent.md"
  },
  "agentAssist": {
    "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
    "steps": [
      "Download the package from Yavira.",
      "Extract it into a folder your agent can access.",
      "Paste one of the prompts below and point your agent at the extracted folder."
    ],
    "prompts": [
      {
        "label": "New install",
        "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
      },
      {
        "label": "Upgrade existing",
        "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
      }
    ]
  },
  "documentation": {
    "source": "clawhub",
    "primaryDoc": "SKILL.md",
    "sections": [
      {
        "title": "Anima Avatar (Project Anima)",
        "body": "Generates high-quality interactive videos where Shutiao speaks the text with appropriate expressions, gestures, and voice."
      },
      {
        "title": "Capabilities",
        "body": "True Voice: Uses Fish Audio API for realistic speech synthesis.\nDynamic Sprites: Auto-selects from a library of 30+ sprites (Happy, Angry, Shy, Think, Action) based on emotion tags.\nSmart Director: Handles parallel rendering, audio-sync, and video composition (FFmpeg).\nPro Delivery: Uploads as native stream to Feishu for direct playback (with correct duration)."
      },
      {
        "title": "Structure",
        "body": "src/director.js: The core engine. Generates frames (sharp + SVG), audio (Fish Audio), and video (FFmpeg).\nsrc/send_video_pro.js: Delivery script. Handles transcoding, duration calculation, and Feishu upload.\nsrc/batch_generator.js: Batch sprite generator. Uses Gemini image generation to produce sprite variants.\nassets/sprites/: The sprite library (1920x1080 PNG files).\nassets/production_plan.csv: The asset registry (25 sprites).\nassets/manifest.json: Sprite metadata for reference.\noutput/: Generated videos."
      },
      {
        "title": "IMPORTANT: Sprites Not Included",
        "body": "ClawHub only distributes text files. The sprite PNG images are not included in the published package.\n\nAfter installing, follow the steps below in order to prepare your sprites before first use.\n\nAll image generation steps use Gemini API (Nano Banana) as the AI image generator. It works by \"reference image + text prompt\" — you give it an existing image and a text description of what to change, and it returns a new image with the changes applied. This is how both the base sprite (character + background fusion) and all expression variants are created."
      },
      {
        "title": "Step 1: Prepare your character image",
        "body": "You need a standalone character illustration (transparent background PNG recommended).\n\nThis is your character's \"identity\" — it defines the look for all sprites.\nResolution: at least 1920x1080. Full-body is best.\nExample: a full-body anime character PNG with transparent background.\n\nSave it somewhere accessible (e.g. avatars/my_character.png)."
      },
      {
        "title": "Step 2: Prepare your background image",
        "body": "You need a background scene for the character to stand in.\n\nThis is the environment that appears behind the character in every video frame.\nResolution: at least 1920x1080.\nExample: a cherry blossom garden, a classroom, a city street.\n\nSave it at: assets/backgrounds/ (e.g. assets/backgrounds/cherry_blossom_bg.png)."
      },
      {
        "title": "Step 3: Fuse character + background into base sprite",
        "body": "This step uses Gemini (Nano Banana) image generation to merge your character onto the background. The AI sees both images and creates a natural-looking composite — this is NOT a simple overlay/paste, but an AI-generated fusion that handles lighting, shadows, and blending.\n\nHow to do it:\n\nMethod A: Use Gemini directly (recommended)\nUse any Gemini-compatible image generation tool (like Nano Banana, Google AI Studio, or the Gemini API) with:\n\nInput image: Your background image\nReference/overlay: Your character image\nPrompt: e.g. \"Place this character naturally in the center of this background scene, full body visible, gentle smile\"\n\nSave the output as: assets/sprites/shutiao_base.png\n\nMethod B: Use the built-in compose script (simple overlay)\nIf you just want a quick mechanical overlay (no AI blending), src/compose_base.js can paste your character onto the background using sharp:\n\nEdit src/compose_base.js — update BG_PATH and AVATAR_PATH to point to your files.\nRun: node src/compose_base.js\nOutput: assets/sprites/shutiao_base.png\n\nNote: Method B is a plain image composite. Method A (Gemini) produces much better results because it handles lighting and integration naturally."
      },
      {
        "title": "Step 4: Plan your sprite variants",
        "body": "Now that you have a base sprite, plan what expression/pose variants you want.\n\nOpen assets/production_plan.csv and customize it:\n\nID,Emotion,Variant,Description,Filename,Prompt,Status\n001,Base,v1,Standard,shutiao_base.png,gentle smile looking at viewer,Done\n003,Happy,v1,Smile,shutiao_happy.png,big happy smile eyes closed,Pending\n007,Angry,v1,Pout,shutiao_angry.png,angry face pouting,Pending\n...\n\nColumn meanings:\n\nEmotion: Category used by the video director to pick sprites (Happy, Angry, Shy, Think, Sad, Action, Base).\nFilename: Output filename. Must follow shutiao_<emotion>_<variant>.png format.\nPrompt: Describes how this variant differs from the base. The generator sends the base image + this prompt to Gemini, asking it to change only the expression/pose while keeping everything else the same.\nStatus: Pending = will be generated. Done = already exists, skip.\n\nThe default CSV has 25 entries. You can add, remove, or modify rows freely."
      },
      {
        "title": "Step 5: Generate sprite variants",
        "body": "This step uses Gemini (Nano Banana) image generation again. For each Pending row, the batch generator sends your base sprite + the prompt to Gemini, asking: \"Same image, change facial expression to [prompt]. Keep clothes and background exactly same.\"\n\nSet your Gemini API key in skills/anima/.env:\n\nGEMINI_API_KEY=your_key_here\n\nMake sure assets/sprites/shutiao_base.png (or shutiao_base_1k.png) exists from Step 3.\n\n\nRun the batch generator:\n\nnode skills/anima/src/batch_generator.js\n\nWhat happens:\n\nReads production_plan.csv\nFinds all rows with Status=Pending\nFor each: sends the base sprite + prompt to Gemini API\nSaves the generated image as a PNG in assets/sprites/\nUpdates the CSV row to Status=Done\nWaits 10 seconds between generations (API rate limit cooldown)"
      },
      {
        "title": "Step 6: Verify",
        "body": "Check that assets/sprites/ now has a PNG file for every row in production_plan.csv:\n\nls assets/sprites/*.png | wc -l\n\nThen do a quick test run:\n\nnode skills/anima/run.js --preview --script '[{\"text\":\"Test\",\"emotion\":\"Happy\"}]'\n\nCheck the generated frame at temp/frame_0.png — you should see your character with the text overlay.\n\nIf a sprite is missing at runtime, the director will fall back to a white background with a warning in the console."
      },
      {
        "title": "1. System Dependencies",
        "body": "ffmpeg (required for video processing):\n\nmacOS: brew install ffmpeg\nLinux: sudo apt install ffmpeg\nWindows: Download/Install FFmpeg and add to PATH."
      },
      {
        "title": "2. Node Dependencies",
        "body": "Install inside the skill folder:\n\ncd skills/anima\nnpm install\n\nThe only native dependency is sharp, which ships prebuilt binaries for all major platforms via N-API. It does not need recompilation when Node versions change — install once, run everywhere."
      },
      {
        "title": "3. External Services (API Keys Required)",
        "body": "This skill depends on two external services. You need to provide your own API keys.\n\nFish Audio (TTS - Text to Speech)\n\nWhat: Generates realistic voice audio from text.\nUsed by: src/director.js (the generateAudio() function).\nGet a key: https://fish.audio/dashboard/api\nEnv vars needed:\n\nFISH_AUDIO_KEY — Your API key (starts with sk-... or a hex string).\nFISH_AUDIO_REF_ID — The voice model reference ID. You can use Fish Audio's default models or clone your own voice.\n\nGemini API (Image Generation - Optional)\n\nWhat: Generates sprite variants using Google Gemini image generation.\nUsed by: src/batch_generator.js (only needed if you want to create new sprite variants).\nSelf-contained: No external skills needed. batch_generator.js calls the Gemini API directly via curl.\nGet a key: https://aistudio.google.com/apikey\nEnv var needed: GEMINI_API_KEY\nNot needed for normal video generation — only for creating new character sprites.\n\nFeishu / Lark (Delivery - Optional)\n\nWhat: Uploads videos to Feishu as native media messages.\nUsed by: src/send_video_pro.js.\nEnv vars needed:\n\nFEISHU_APP_ID — Your Feishu app ID.\nFEISHU_APP_SECRET — Your Feishu app secret.\n\n\nNot needed if you only use --preview mode."
      },
      {
        "title": "4. Environment Configuration",
        "body": "Create a .env file inside the skill folder (skills/anima/.env):\n\n# Fish Audio (Required for TTS)\nFISH_AUDIO_KEY=your_key_here\nFISH_AUDIO_REF_ID=your_model_ref_id_here\n\n# Gemini (Optional, for sprite generation)\nGEMINI_API_KEY=your_key_here\n\n# Feishu/Lark (Optional, for delivery)\nFEISHU_APP_ID=cli_...\nFEISHU_APP_SECRET=...\n\nImportant: The .env file is loaded from the skill folder first (least-privilege). Never commit .env files — the .clawignore already excludes it."
      },
      {
        "title": "Generate & Send",
        "body": "# Basic usage (Demo script)\nnode skills/anima/run.js --target \"ou_...\"\n\n# With custom script (JSON string)\nnode skills/anima/run.js --target \"ou_...\" --script '[{\"text\":\"Hello World\",\"emotion\":\"Happy\"}]'\n\n# With custom script (File)\nnode skills/anima/run.js --target \"ou_...\" --script \"path/to/script.json\"\n\n# Preview only (No upload)\nnode skills/anima/run.js --script '[{\"text\":\"Test\",\"emotion\":\"Happy\"}]' --preview"
      },
      {
        "title": "One-Liner (for agent use)",
        "body": "node skills/anima/run.js --target \"<open_id>\" --script '[{\"text\":\"Hello\",\"emotion\":\"Happy\"}]'"
      },
      {
        "title": "Script Format",
        "body": "Each scene in the script is a JSON object:\n\n[\n  { \"text\": \"Hello boss!\", \"emotion\": \"Happy\" },\n  { \"text\": \"Let me think...\", \"emotion\": \"Think\" },\n  { \"text\": \"I got it!\", \"emotion\": \"Action\" }\n]\n\nAvailable emotions: Base, Happy, Angry, Shy, Think, Sad, Action."
      },
      {
        "title": "Extension: Custom TTS",
        "body": "To use a different TTS provider (e.g., OpenAI, ElevenLabs):\n\nOpen src/director.js.\nLocate the generateAudio(text, filename) function.\nReplace the Fish Audio API call with your provider's logic.\nContract: The function must return: { path: \"/path/to/audio.wav\", duration: 1.5 } (duration in seconds)."
      },
      {
        "title": "Advanced: Adding More Sprite Variants",
        "body": "To add new expressions or poses after the initial setup:\n\nAdd a new row to assets/production_plan.csv with Status=Pending.\nWrite a clear prompt describing the change from the base (e.g. angry expression, arms crossed, looking away).\nRun node src/batch_generator.js — it will only process Pending rows.\nThe new sprite will auto-register in the director's emotion pool via loadSprites().\n\nSee ASSETS_PLAN.md for the full production matrix and design philosophy."
      },
      {
        "title": "Troubleshooting",
        "body": "Duration 00:00: Ensure send_video_pro.js calculates duration in ms and passes it to both upload and message payload.\nFish Audio 400: Check that your Ref ID matches the API Key owner's model.\nVideo Black: Check ffmpeg transcoding logs and verify source frame images in temp/frame_*.png.\nSVG text not rendering: Ensure the system has CJK fonts installed (macOS has them by default; on Linux: sudo apt install fonts-noto-cjk).\nNo audio fallback: If FISH_AUDIO_KEY is missing, the skill falls back to macOS say command (English only)."
      }
    ],
    "body": "Anima Avatar (Project Anima)\n\nGenerates high-quality interactive videos where Shutiao speaks the text with appropriate expressions, gestures, and voice.\n\nCapabilities\nTrue Voice: Uses Fish Audio API for realistic speech synthesis.\nDynamic Sprites: Auto-selects from a library of 30+ sprites (Happy, Angry, Shy, Think, Action) based on emotion tags.\nSmart Director: Handles parallel rendering, audio-sync, and video composition (FFmpeg).\nPro Delivery: Uploads as native stream to Feishu for direct playback (with correct duration).\nStructure\nsrc/director.js: The core engine. Generates frames (sharp + SVG), audio (Fish Audio), and video (FFmpeg).\nsrc/send_video_pro.js: Delivery script. Handles transcoding, duration calculation, and Feishu upload.\nsrc/batch_generator.js: Batch sprite generator. Uses Gemini image generation to produce sprite variants.\nassets/sprites/: The sprite library (1920x1080 PNG files).\nassets/production_plan.csv: The asset registry (25 sprites).\nassets/manifest.json: Sprite metadata for reference.\noutput/: Generated videos.\nIMPORTANT: Sprites Not Included\n\nClawHub only distributes text files. The sprite PNG images are not included in the published package.\n\nAfter installing, follow the steps below in order to prepare your sprites before first use.\n\nAll image generation steps use Gemini API (Nano Banana) as the AI image generator. It works by \"reference image + text prompt\" — you give it an existing image and a text description of what to change, and it returns a new image with the changes applied. This is how both the base sprite (character + background fusion) and all expression variants are created.\n\nStep 1: Prepare your character image\n\nYou need a standalone character illustration (transparent background PNG recommended).\n\nThis is your character's \"identity\" — it defines the look for all sprites.\nResolution: at least 1920x1080. Full-body is best.\nExample: a full-body anime character PNG with transparent background.\n\nSave it somewhere accessible (e.g. avatars/my_character.png).\n\nStep 2: Prepare your background image\n\nYou need a background scene for the character to stand in.\n\nThis is the environment that appears behind the character in every video frame.\nResolution: at least 1920x1080.\nExample: a cherry blossom garden, a classroom, a city street.\n\nSave it at: assets/backgrounds/ (e.g. assets/backgrounds/cherry_blossom_bg.png).\n\nStep 3: Fuse character + background into base sprite\n\nThis step uses Gemini (Nano Banana) image generation to merge your character onto the background. The AI sees both images and creates a natural-looking composite — this is NOT a simple overlay/paste, but an AI-generated fusion that handles lighting, shadows, and blending.\n\nHow to do it:\n\nMethod A: Use Gemini directly (recommended) Use any Gemini-compatible image generation tool (like Nano Banana, Google AI Studio, or the Gemini API) with:\n\nInput image: Your background image\nReference/overlay: Your character image\nPrompt: e.g. \"Place this character naturally in the center of this background scene, full body visible, gentle smile\"\n\nSave the output as: assets/sprites/shutiao_base.png\n\nMethod B: Use the built-in compose script (simple overlay) If you just want a quick mechanical overlay (no AI blending), src/compose_base.js can paste your character onto the background using sharp:\n\nEdit src/compose_base.js — update BG_PATH and AVATAR_PATH to point to your files.\nRun: node src/compose_base.js\nOutput: assets/sprites/shutiao_base.png\n\nNote: Method B is a plain image composite. Method A (Gemini) produces much better results because it handles lighting and integration naturally.\n\nStep 4: Plan your sprite variants\n\nNow that you have a base sprite, plan what expression/pose variants you want.\n\nOpen assets/production_plan.csv and customize it:\n\nID,Emotion,Variant,Description,Filename,Prompt,Status\n001,Base,v1,Standard,shutiao_base.png,gentle smile looking at viewer,Done\n003,Happy,v1,Smile,shutiao_happy.png,big happy smile eyes closed,Pending\n007,Angry,v1,Pout,shutiao_angry.png,angry face pouting,Pending\n...\n\n\nColumn meanings:\n\nEmotion: Category used by the video director to pick sprites (Happy, Angry, Shy, Think, Sad, Action, Base).\nFilename: Output filename. Must follow shutiao_<emotion>_<variant>.png format.\nPrompt: Describes how this variant differs from the base. The generator sends the base image + this prompt to Gemini, asking it to change only the expression/pose while keeping everything else the same.\nStatus: Pending = will be generated. Done = already exists, skip.\n\nThe default CSV has 25 entries. You can add, remove, or modify rows freely.\n\nStep 5: Generate sprite variants\n\nThis step uses Gemini (Nano Banana) image generation again. For each Pending row, the batch generator sends your base sprite + the prompt to Gemini, asking: \"Same image, change facial expression to [prompt]. Keep clothes and background exactly same.\"\n\nSet your Gemini API key in skills/anima/.env:\nGEMINI_API_KEY=your_key_here\n\n\nMake sure assets/sprites/shutiao_base.png (or shutiao_base_1k.png) exists from Step 3.\n\nRun the batch generator:\n\nnode skills/anima/src/batch_generator.js\n\n\nWhat happens:\n\nReads production_plan.csv\nFinds all rows with Status=Pending\nFor each: sends the base sprite + prompt to Gemini API\nSaves the generated image as a PNG in assets/sprites/\nUpdates the CSV row to Status=Done\nWaits 10 seconds between generations (API rate limit cooldown)\nStep 6: Verify\n\nCheck that assets/sprites/ now has a PNG file for every row in production_plan.csv:\n\nls assets/sprites/*.png | wc -l\n\n\nThen do a quick test run:\n\nnode skills/anima/run.js --preview --script '[{\"text\":\"Test\",\"emotion\":\"Happy\"}]'\n\n\nCheck the generated frame at temp/frame_0.png — you should see your character with the text overlay.\n\nIf a sprite is missing at runtime, the director will fall back to a white background with a warning in the console.\n\nSetup & Requirements\n1. System Dependencies\nffmpeg (required for video processing):\nmacOS: brew install ffmpeg\nLinux: sudo apt install ffmpeg\nWindows: Download/Install FFmpeg and add to PATH.\n2. Node Dependencies\n\nInstall inside the skill folder:\n\ncd skills/anima\nnpm install\n\n\nThe only native dependency is sharp, which ships prebuilt binaries for all major platforms via N-API. It does not need recompilation when Node versions change — install once, run everywhere.\n\n3. External Services (API Keys Required)\n\nThis skill depends on two external services. You need to provide your own API keys.\n\nFish Audio (TTS - Text to Speech)\nWhat: Generates realistic voice audio from text.\nUsed by: src/director.js (the generateAudio() function).\nGet a key: https://fish.audio/dashboard/api\nEnv vars needed:\nFISH_AUDIO_KEY — Your API key (starts with sk-... or a hex string).\nFISH_AUDIO_REF_ID — The voice model reference ID. You can use Fish Audio's default models or clone your own voice.\nGemini API (Image Generation - Optional)\nWhat: Generates sprite variants using Google Gemini image generation.\nUsed by: src/batch_generator.js (only needed if you want to create new sprite variants).\nSelf-contained: No external skills needed. batch_generator.js calls the Gemini API directly via curl.\nGet a key: https://aistudio.google.com/apikey\nEnv var needed: GEMINI_API_KEY\nNot needed for normal video generation — only for creating new character sprites.\nFeishu / Lark (Delivery - Optional)\nWhat: Uploads videos to Feishu as native media messages.\nUsed by: src/send_video_pro.js.\nEnv vars needed:\nFEISHU_APP_ID — Your Feishu app ID.\nFEISHU_APP_SECRET — Your Feishu app secret.\nNot needed if you only use --preview mode.\n4. Environment Configuration\n\nCreate a .env file inside the skill folder (skills/anima/.env):\n\n# Fish Audio (Required for TTS)\nFISH_AUDIO_KEY=your_key_here\nFISH_AUDIO_REF_ID=your_model_ref_id_here\n\n# Gemini (Optional, for sprite generation)\nGEMINI_API_KEY=your_key_here\n\n# Feishu/Lark (Optional, for delivery)\nFEISHU_APP_ID=cli_...\nFEISHU_APP_SECRET=...\n\n\nImportant: The .env file is loaded from the skill folder first (least-privilege). Never commit .env files — the .clawignore already excludes it.\n\nUsage\nGenerate & Send\n# Basic usage (Demo script)\nnode skills/anima/run.js --target \"ou_...\"\n\n# With custom script (JSON string)\nnode skills/anima/run.js --target \"ou_...\" --script '[{\"text\":\"Hello World\",\"emotion\":\"Happy\"}]'\n\n# With custom script (File)\nnode skills/anima/run.js --target \"ou_...\" --script \"path/to/script.json\"\n\n# Preview only (No upload)\nnode skills/anima/run.js --script '[{\"text\":\"Test\",\"emotion\":\"Happy\"}]' --preview\n\nOne-Liner (for agent use)\nnode skills/anima/run.js --target \"<open_id>\" --script '[{\"text\":\"Hello\",\"emotion\":\"Happy\"}]'\n\nScript Format\n\nEach scene in the script is a JSON object:\n\n[\n  { \"text\": \"Hello boss!\", \"emotion\": \"Happy\" },\n  { \"text\": \"Let me think...\", \"emotion\": \"Think\" },\n  { \"text\": \"I got it!\", \"emotion\": \"Action\" }\n]\n\n\nAvailable emotions: Base, Happy, Angry, Shy, Think, Sad, Action.\n\nExtension: Custom TTS\n\nTo use a different TTS provider (e.g., OpenAI, ElevenLabs):\n\nOpen src/director.js.\nLocate the generateAudio(text, filename) function.\nReplace the Fish Audio API call with your provider's logic.\nContract: The function must return: { path: \"/path/to/audio.wav\", duration: 1.5 } (duration in seconds).\nAdvanced: Adding More Sprite Variants\n\nTo add new expressions or poses after the initial setup:\n\nAdd a new row to assets/production_plan.csv with Status=Pending.\nWrite a clear prompt describing the change from the base (e.g. angry expression, arms crossed, looking away).\nRun node src/batch_generator.js — it will only process Pending rows.\nThe new sprite will auto-register in the director's emotion pool via loadSprites().\n\nSee ASSETS_PLAN.md for the full production matrix and design philosophy.\n\nTroubleshooting\nDuration 00:00: Ensure send_video_pro.js calculates duration in ms and passes it to both upload and message payload.\nFish Audio 400: Check that your Ref ID matches the API Key owner's model.\nVideo Black: Check ffmpeg transcoding logs and verify source frame images in temp/frame_*.png.\nSVG text not rendering: Ensure the system has CJK fonts installed (macOS has them by default; on Linux: sudo apt install fonts-noto-cjk).\nNo audio fallback: If FISH_AUDIO_KEY is missing, the skill falls back to macOS say command (English only)."
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/HMyaoyuan/anima",
    "publisherUrl": "https://clawhub.ai/HMyaoyuan/anima",
    "owner": "HMyaoyuan",
    "version": "3.3.2",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/anima",
    "downloadUrl": "https://openagent3.xyz/downloads/anima",
    "agentUrl": "https://openagent3.xyz/skills/anima/agent",
    "manifestUrl": "https://openagent3.xyz/skills/anima/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/anima/agent.md"
  }
}