[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6MNQ4XN9TXF8Y2HGACE9X2V":3,"skill-same:01M6MNQ4XN9TXF8Y2HGACE9X2V":87},{"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":26,"publishedAt":47,"linkOnly":48,"markdownContent":49,"markdownBody":50,"frontMatter":51,"refName":63,"commitSha":64,"skillDirectory":65,"forksCount":66,"repositoryHtmlUrl":67,"repositoryHomepageUrl":68,"ownerHtmlUrl":69,"ownerBlogUrl":70,"ownerDescription":71,"updatedAt":72,"install":73,"resources":86},"01M6MNQ4XN9TXF8Y2HGACE9X2V","varg-consistent-characters",null,"Makes multi-scene ad videos on varg in which the same people, outfits and location stay consistent in every shot, built from real actor photos or generated characters: one profile card per actor, an empty location card, scene composites approved as stills first, then image-to-video with native dialogue audio, trimmed and captioned in one cloud render. Also covers frame and audio QA of the output and turning a reference video into a scene script. Use when a script names actors (A1, A2), a location (L1) and numbered scenes, when someone sends actor or model photos, or asks for a consistent character, the same person in every scene, a recurring mascot or brand character, a UGC or skit ad with two actors, or to remake a reference ad with new people. Not for one speaker on a stage (use varg-character-keynote), seated two-person podcast dialogue (use varg-podcast-video), music videos (use varg-music-video), or one-off images and clips (use varg-ai).","hashgraph-online\u002Fawesome-codex-plugins","hashgraph-online","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F160371962?v=4","01M6CNB6NRJVNNGN14TR01Q6D3","HOL (Hashgraph Online)","01M6JT1HSKKG9V68KQAQGBA7XP",1575,"MIT","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002FvargHQ\u002Fskills\u002Fvarg-consistent-characters\u002FSKILL.md",[18,23],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000016","图像与多媒体","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",[27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46],"varg consistent characters","characters","character","frame","for","which","image-to-video","composites","card","as","MCP","use","edit","write","read","brand","music","Audio","Video","image","2026-11-10 10:12:30",false,"---\nname: varg-consistent-characters\ndescription: >-\n  Makes multi-scene ad videos on varg in which the same people, outfits and location stay\n  consistent in every shot, built from real actor photos or generated characters: one profile\n  card per actor, an empty location card, scene composites approved as stills first, then\n  image-to-video with native dialogue audio, trimmed and captioned in one cloud render. Also\n  covers frame and audio QA of the output and turning a reference video into a scene script.\n  Use when a script names actors (A1, A2), a location (L1) and numbered scenes, when someone\n  sends actor or model photos, or asks for a consistent character, the same person in every\n  scene, a recurring mascot or brand character, a UGC or skit ad with two actors, or to remake a\n  reference ad with new people. Not for one speaker on a stage (use varg-character-keynote),\n  seated two-person podcast dialogue (use varg-podcast-video), music videos (use\n  varg-music-video), or one-off images and clips (use varg-ai).\nlicense: MIT\nmetadata:\n  author: vargHQ\n  version: \"1.0.0\"\n  openclaw:\n    requires:\n      anyBins:\n        - curl\n        - bun\n    homepage: https:\u002F\u002Fvarg.ai\ncompatibility: >-\n  Needs a VARG_API_KEY (see the varg-ai skill) or the varg MCP connection\n  (https:\u002F\u002Fmcp.varg.ai\u002Fmcp). The images-only approval script needs bun; the final render runs in\n  the cloud with curl. QA uses ffmpeg\u002Fffprobe locally.\nallowed-tools: Bash(bun:*) Bash(bunx:*) Bash(curl:*) Bash(jq:*) Bash(ffmpeg:*) Bash(ffprobe:*) Bash(nohup:*) Read Write Edit\n---\n\n# varg consistent characters\n\nTurn a scene script + actor photos into a multi-scene vertical ad where characters, outfits and location stay consistent across every cut. Tested on a 5-scene, 22-second 9:16 fitness ad with two real actors in one gym.\n\nGeneration runs on varg: the `varg-ai` skill (sign-in, models, local and cloud render) or the varg MCP server (`https:\u002F\u002Fmcp.varg.ai\u002Fmcp`). This skill assumes a working `VARG_API_KEY` or MCP connection.\n\n## Core idea: cards, then composites\n\nNever describe characters from scratch in every scene prompt. Build a small set of **frozen reference images** once, then compose every scene from them.\n\n```\nphoto A1 ──edit──> a1Card ─┐\nphoto A2 ──edit──> a2Card ─┼──edit──> scene1..N ──image-to-video──> vid1..N\nprompt   ──gen───> gymLoc ─┘\n```\n\nUpload the actor photos first (`POST \u002Fv2\u002Ffiles`, or the MCP upload tools) and use the returned URLs.\n\nNo actor photos? Design the cast with `varg-character-design` (archetypes, a model sheet per character on flat grey) and use each character's full-body crop as the profile card below.\n\n## 1. Consistency\n\n### 1.1 Character profile cards\n\nOne image-edit call per actor, from their photo:\n\n```tsx\nconst a1Card = Image({\n  prompt: {\n    text: \"Full-body character reference card of the exact same woman from the photo: same face, same grey hair, same body shape. \u003Cage\u002Fbuild>. She wears \u003CFIXED OUTFIT, described precisely>. Standing straight facing camera, neutral relaxed pose, arms at her sides. Plain light grey studio background, even soft lighting, head-to-toe framing, photorealistic\",\n    images: [A1_PHOTO_URL],\n  },\n  model: varg.imageModel(\"nano_banana_2_1\"),   \u002F\u002F edits when images are passed\n  aspectRatio: \"9:16\",\n});\n```\n\nRules that make cards work:\n\n- **\"exact same ... from the photo: same face, same hair, same body shape\"** anchors identity to the reference instead of the text description.\n- **Fix the outfit in the card.** The card's clothes become the outfit for the whole video. Pick it so the script works (e.g. arms or legs visible if scenes point at them). Never re-describe clothes differently later.\n- **Full body, neutral pose, plain studio background**: a clean card composites into any scene; a scene-styled card does not.\n\nThe pipeline was first built with Nano Banana Pro edit; `nano_banana_2_1` is its successor and varg's default. For the hardest identity cases try `gpt_image_2_5`.\n\n### 1.2 Location card\n\nOne text-to-image generation, **no people**:\n\n```tsx\nconst gymLoc = Image({\n  prompt: \"Empty modern gym interior with huge floor-to-ceiling windows... No people. ... photorealistic, 9:16 portrait framing\",\n  model: varg.imageModel(\"nano_banana_2_1\"),\n  aspectRatio: \"9:16\",\n});\n```\n\nAn empty location card lets every scene reuse the same walls, light and floor without characters from a previous scene leaking in.\n\n### 1.3 Scene composites\n\nEvery scene is one edit call with **all three refs**:\n\n```tsx\nconst SAME =\n  \"First reference image is Woman A (\u003Cshort recap of card>). Second reference image is Woman B (\u003Cshort recap>). Third reference image is the gym location with huge windows. Keep both women's faces, hair, bodies and outfits exactly as in their reference cards, and the gym exactly as in the location reference.\";\n\nconst scene2 = Image({\n  prompt: { text: `${SAME} \u003Cscene-specific framing, poses, gestures>`, images: [a1Card, a2Card, gymLoc] },\n  model: varg.imageModel(\"nano_banana_2_1\"),\n  aspectRatio: \"9:16\",\n});\n```\n\n- The `SAME` preamble is **shared verbatim** across all scenes. It names each reference by position (\"First reference image is...\") and pins identity.\n- Image edit models take several refs fine. For **Video**, one image = the first frame; several images switch to reference mode, which only some models support (see `varg-ai`).\n- Scene text only describes what changes: camera framing, poses, gestures, expressions.\n\n### 1.4 Cache discipline (critical)\n\nIdentical prompt + params = free cache hit, both locally and **server-side on api.varg.ai**. This is what makes the whole pipeline cheap:\n\n- Keep prompts **byte-identical** between the images-only script, the local template and the cloud template. Copy-paste, never retype.\n- The full 5-scene video re-rendered in the cloud for a few credits (the stitching fee) because every image, video and music prompt matched the earlier runs.\n- Editing one scene's prompt regenerates only that scene; everything else stays cached.\n\n### 1.5 Content-moderation gotcha\n\nBody-part wording such as \"rear\" or \"booty\" in an **image** prompt can silently stall generation for ~10 min (provider moderation). Rephrase to neutral anatomy: \"hips and thighs\". The same words inside spoken dialogue in **video** prompts passed fine.\n\nFitness and swimwear scenes trip filters on words, not pictures: \"girl\" instead of \"woman, 28\", \"tight\", \"bikini\" and short prompts without a garment description get blocked on GPT Image 2.5, FLUX 3 and Grok, while a catalog-style prompt (adult age as a number, each garment named, an activity, \"activewear brand catalog photo\") passed on all six image models we tried. A block can be random: retry once, then rephrase, then route. Test log, safe-phrasing table and routing: `varg-character-design` → `references\u002Fmoderation.md`.\n\n## 2. Photos-for-approval pass (images first)\n\nGenerate and show all stills **before** paying for any video. A still costs a few credits; a video clip costs ten times more or more. Approve the cheap layer first.\n\nStandalone script pattern (`\u003Cproject>-images.ts`, run with bun):\n\n```ts\nimport { Image } from \"vargai\u002Freact\";\nimport { createVarg } from \"vargai\u002Fai\";\n\nconst varg = createVarg({ apiKey: process.env.VARG_API_KEY! });\n\u002F\u002F ... define a1Card, a2Card, gymLoc, scene1..sceneN exactly as in 1.x ...\n\nasync function save(name: string, el: ReturnType\u003Ctypeof Image>) {\n  const resolved = await el;               \u002F\u002F await triggers generation (cached)\n  await Bun.write(`output\u002F\u003Cproject>\u002F${name}.png`, await resolved.file.data());\n}\n\n\u002F\u002F cards + location first, then scenes (scenes reference the cards)\nawait Promise.all([save(\"a1-card\", a1Card), save(\"a2-card\", a2Card), save(\"location\", gymLoc)]);\nawait Promise.all([save(\"scene1\", scene1), \u002F* ... *\u002F]);\n```\n\nKey facts:\n\n- `Image({...})` is a lazy thenable: `await` resolves it, `.file.data()` gives the bytes. Awaited results are cached, so the later video render reuses them for free.\n- Run it in the background (`nohup bun run \u003Cproject>-images.ts > images.log 2>&1 &`) and report the PID.\n- Show the user: the cards + location + every scene. Iterate on individual scene prompts (only the changed one regenerates) until approved.\n\nNo bun? Generate the same stills one by one with `POST \u002Fv2\u002Fimage` (or MCP `generate_image`), passing the references in `files`. Keep the prompts identical to the ones you will use in the render.\n\n## 3. Final render: cloud (`POST \u002Fv2\u002Frender`)\n\nAfter the stills are approved, wrap them in videos + composition and submit the whole TSX to the cloud. No bun or ffmpeg needed on the machine that submits.\n\n### 3.1 Video layer: image-to-video with native audio\n\nDialogue is written into the video prompt and the model voices it:\n\n```tsx\nconst vid2 = Video({\n  prompt: {\n    text: `Same gym with huge windows. \u003Cscene action, static camera>.\nThe trainer says matter-of-factly: \"\u003Cthe scene's line>\"\nSoft gym ambient sounds.`,\n    images: [scene2],                      \u002F\u002F exactly ONE image = the first frame\n  },\n  model: varg.videoModel(\"kling_v4\"),\n  audio: \"native\",                         \u002F\u002F generates and keeps the dialogue audio\n  duration: 4,                             \u002F\u002F integer seconds (kling_v4: 3-30)\n});\n```\n\n- Quote dialogue lines verbatim in the prompt; add a voice direction (\"clear encouraging voice\", \"matter-of-factly\").\n- Budget ~1 s of dialogue per ~2.5 words, then trim (3.3).\n- `kling_v3` is what this pipeline was first tested on and still works the same way (integer 3-15 s); `kling_v4` is the current Kling. `seedance_2_5` gives the best faces and lip movement at a higher price.\n- Native audio needs `audio: \"native\"`. A flat `providerOptions: { varg: { generate_audio: true } }` is silently dropped and you get a silent clip.\n\n### 3.2 Cloud TSX rules\n\n`POST https:\u002F\u002Fapi.varg.ai\u002Fv2\u002Frender` with `{\"code\": \"\u003Ctsx>\"}` (MCP: `render_video` with `code`):\n\n- **No imports, no `createVarg`**: `Render\u002FClip\u002FImage\u002FVideo\u002FMusic\u002FCaptions` and `varg` are injected as globals. Local LSP errors on this file are expected and fine.\n- Must `export default (\u003CRender ...>...)`; no named exports.\n- Same `varg.*Model()` ids as local.\n\nPrice it first (MCP `estimate_cost` with tool `render` and the code, or `POST \u002Fv2\u002Festimate` on each clip's request) and get the user's go-ahead. Then submit and poll:\n\n```bash\ncurl -s -X POST https:\u002F\u002Fapi.varg.ai\u002Fv2\u002Frender \\\n  -H \"Authorization: Bearer $VARG_API_KEY\" -H \"Content-Type: application\u002Fjson\" \\\n  -d \"{\\\"code\\\": $(jq -Rs . \u003C template-cloud.tsx)}\"\n# -> {\"id\":\"job_xxx\",...}; poll every 15s:\ncurl -s https:\u002F\u002Fapi.varg.ai\u002Fv2\u002Fjobs\u002FJOB_ID -H \"Authorization: Bearer $VARG_API_KEY\"\n# completed -> output.outputs[0].url  (s3.varg.ai mp4)\n```\n\nIf all prompts match a previous run, the cloud render costs a few credits (server cache) and finishes in ~2 min.\n\n### 3.3 Dynamic pacing: trim dead air\n\nGenerated clips have ~0.2-0.5 s of warm-up and often trailing silence. Detect speech boundaries on the generated clips, then cut tight:\n\n```bash\nffmpeg -i scene.mp4 -af \"silencedetect=noise=-35dB:d=0.35\" -f null -\n```\n\n```tsx\n\u003CClip cutFrom={0.2} cutTo={5.6} duration={5.4}>{vid1}\u003C\u002FClip>  \u002F\u002F cut right after the line ends\n```\n\nThis took the test video from 27 s to 22.4 s with no content loss.\nKeep `Music` `duration` an **integer**: a fractional total like 22.4 failed with a 422 on ElevenLabs music. Round down.\n\n### 3.4 Centered captions in the cloud (inline SRT)\n\n`\u003CCaptions>` transcribes `Speech` elements by itself. For dialogue that lives in a video's native audio, the tested route is two passes (newer vargai builds also accept `src={vid.audio}`; if the render rejects that with \"Captions src must be ...\", use this route):\n\n1. Render once without captions (cloud or local).\n2. Upload the audio and transcribe it on varg: `POST \u002Fv2\u002Ffiles` (upload) -> `POST \u002Fv2\u002Ftranscription` with `{\"model\":\"whisper_large_v3\",\"audio_url\":...}`. The result has word-level timestamps.\n3. Build a word-level SRT; fix Whisper mishears (product names, unusual words). Captions' `prompt` prop can also pass context hints to Whisper.\n4. Embed the SRT **inline** in the cloud TSX. The `srt` prop takes a file path, and template code runs on the render server, so write the file to `\u002Ftmp` at module eval:\n\n```tsx\nconst SRT_CONTENT = `1\n00:00:00,260 --> 00:00:00,320\nYou\n...`;\nconst SRT_PATH = `\u002Ftmp\u002Fproject-${Date.now()}.srt`;\nawait Bun.write(SRT_PATH, SRT_CONTENT);\n\u002F\u002F in \u003CRender>:\n\u003CCaptions srt={SRT_PATH} style=\"tiktok\" position=\"center\"\n          color=\"#ffffff\" activeColor=\"#a3e635\" wordsPerLine={3} fontSize={64} \u002F>\n```\n\nRe-submitting is nearly free (everything cached); only the composition re-runs.\nImportant: if you change any clip trim, **regenerate the SRT**, because the timings shift.\n\n## 4. QA: how to analyze what you generated\n\nAgents can't watch video, but they can read frames and measure audio.\n\n### 4.1 Visual QA: extract frames and READ them\n\nAfter every render (and every images pass), pull frames and inspect them as images. Never ship a video you haven't frame-checked.\n\n```bash\n# one frame per scene, at a timestamp INSIDE that scene (mind the trims)\nffmpeg -v error -ss 3  -i out.mp4 -frames:v 1 -vf scale=270:-1 check-s1.png -y\nffmpeg -v error -ss 9  -i out.mp4 -frames:v 1 -vf scale=270:-1 check-s3.png -y\nffmpeg -v error -ss 14 -i out.mp4 -frames:v 1 -vf scale=270:-1 check-s4.png -y\n```\n\n- Downscale (`scale=270:-1` .. `360:-1`): enough to judge, cheap on context.\n- Read several frames **in one parallel tool call**.\n- No local ffmpeg? The varg MCP `analyze_video` tool answers questions about a rendered video.\n\nVisual checklist per frame:\n\n- **Identity**: same face\u002Fhair\u002Fbody as the profile cards, in every scene.\n- **Outfit drift**: clothes must match the cards exactly (color, cut, shoes). Drift usually means a scene prompt re-described clothing: fix the wording, not the card.\n- **Location drift**: windows\u002Ffloor\u002Fprops consistent with the location card.\n- **Framing**: matches the script (full-body vs close-up); subjects not cropped at the wrong place; both actors present when required.\n- **Captions**: correct position (center), readable size, active-word color visible, no overlap with faces.\n- **Age\u002Fbuild drift**: edit models can age or slim a person between scenes. Compare against the card and regenerate the outlier scene only.\n\nAlso verify structure numerically:\n\n```bash\nffprobe -v error -show_entries format=duration,size -of default=noprint_wrappers=1 out.mp4\n```\n\nDuration must equal the sum of clip durations; a mismatch means a clip or music duration bug, not an encoding issue.\n\n### 4.2 Audio QA: silencedetect as ears\n\n`silencedetect` shows where speech actually lives in each generated clip:\n\n```bash\nffmpeg -i scene.mp4 -af \"silencedetect=noise=-35dB:d=0.35\" -f null - 2>&1 | grep silence_\n```\n\nRead it per clip:\n\n- silence at the **start** -> dead air \u002F late speech start -> raise `cutFrom`;\n- silence at the **end** -> line finished early -> lower `cutTo`;\n- **no silence** in an N-second clip -> speech fills it; if the line is long, the model may have rushed or truncated it: listen-check via transcription;\n- silence in the **middle** of a one-line clip -> possible stutter or split delivery.\n\nThen confirm the words: transcribe the final mix (3.4 step 2) and diff the transcript against the script. This catches mispronunciations by the video model and mishears by Whisper. Fix mishears in the SRT; regenerate the clip only if the model actually said the wrong thing.\n\n### 4.3 When to regenerate vs when to trim\n\n| Symptom | Fix |\n|---|---|\n| Dead air, late start, early finish | Trim with `cutFrom`\u002F`cutTo`: free |\n| Wrong\u002Fgarbled spoken line | Regenerate that Video |\n| Identity\u002Foutfit drift in one scene | Regenerate that scene Image, then its Video |\n| Caption timing off after re-trim | Regenerate the SRT only (one transcription) |\n\n## 5. No script? Analyze the reference video first\n\nWhen the user sends a reference video instead of a script, reverse-engineer the same A\u002FL\u002FScene structure this skill consumes. Do this **before** touching any generation. (`media-download` fetches a reference by link.)\n\n### 5.1 Extract structure\n\n```bash\nffprobe -v error -show_entries format=duration -of default=noprint_wrappers=1 ref.mp4\n\n# scene cut candidates (hard cuts)\nffmpeg -i ref.mp4 -vf \"select='gt(scene,0.35)',showinfo\" -f null - 2>&1 | grep showinfo | grep pts_time\n\n# speech map\nffmpeg -i ref.mp4 -af \"silencedetect=noise=-35dB:d=0.35\" -f null - 2>&1 | grep silence_\n```\n\nScene-cut timestamps + speech gaps together give the beat structure: how many scenes, how long each, where dialogue sits.\n\n### 5.2 Sample frames per scene\n\nExtract 1-2 frames inside each detected scene and read them (parallel call):\n\n```bash\nffmpeg -v error -ss \u003Ct> -i ref.mp4 -frames:v 1 -vf scale=360:-1 ref-s1.png -y\n```\n\nFrom the frames, document for each scene:\n\n- **Actors**: who is on screen, rough age\u002Fbuild\u002Fhair -> becomes A1, A2...\n- **Wardrobe**: what each actor wears (this defines the profile-card outfits).\n- **Location**: environment, light, key props -> becomes L1 (L2...).\n- **Framing**: full-body \u002F medium-wide \u002F close-up; front or back view; camera static or moving; who is closer to camera.\n- **Action + gesture**: what each actor does (squats, points at X, thumbs-up).\n\n### 5.3 Get the dialogue\n\nTranscribe the reference audio (same varg call as 3.4). Map each line to its scene via the timestamps. If there's no speech, note the music mood and on-screen text instead.\n\n### 5.4 Write the script, then confirm\n\nAssemble the standard format and **show it to the user for approval** before generating anything:\n\n```\nActors:\n- A1: \u003Cdescription from frames>  - C1: \u003Cwardrobe>\n- A2: \u003Cdescription>\nLocation:\n- L1: \u003Cdescription>\nScene 1 (Ns): \u003Cframing>. \u003Caction>. A2 says: \"\u003Cline>\". L1\nScene 2 ...\n```\n\nFrom here the pipeline is identical: cards (1.1) -> location (1.2) -> composites (1.3) -> approval pass (2) -> videos + cloud render (3). Use the reference's measured scene durations as the initial `duration` values, and its framing notes verbatim in scene prompts.\n\n## Cost reference (5 scenes, ~22 s at 9:16)\n\nRough credits from the varg catalog (October 2026). Always quote the user the exact figure from `POST \u002Fv2\u002Festimate` \u002F `estimate_cost`, in credits.\n\n| Step | Credits (~) |\n|---|---|\n| 8 images (2 cards + location + 5 scenes), `nano_banana_2_1` | ~65 |\n| 5 Kling clips with native audio, 4-6 s at 720p | ~400-500 |\n| Music bed (`mureka_9_5`) + one transcription | ~30 |\n| Cloud re-render with everything cached | a few (stitching fee) |\n| Iterating one scene image \u002F one scene video | ~8 \u002F ~80-100 |\n\n## Related skills\n\n`varg-ai` (sign-in, models, render formats), `varg-character-keynote` (one speaker on a stage), `varg-podcast-video` (two seated characters in dialogue), `varg-previz` (plan many shots and camera moves first), `media-download` (fetch a reference video), `video-skills` (find a skill for another format).\n","# varg consistent characters\n\nTurn a scene script + actor photos into a multi-scene vertical ad where characters, outfits and location stay consistent across every cut. Tested on a 5-scene, 22-second 9:16 fitness ad with two real actors in one gym.\n\nGeneration runs on varg: the `varg-ai` skill (sign-in, models, local and cloud render) or the varg MCP server (`https:\u002F\u002Fmcp.varg.ai\u002Fmcp`). This skill assumes a working `VARG_API_KEY` or MCP connection.\n\n## Core idea: cards, then composites\n\nNever describe characters from scratch in every scene prompt. Build a small set of **frozen reference images** once, then compose every scene from them.\n\n```\nphoto A1 ──edit──> a1Card ─┐\nphoto A2 ──edit──> a2Card ─┼──edit──> scene1..N ──image-to-video──> vid1..N\nprompt   ──gen───> gymLoc ─┘\n```\n\nUpload the actor photos first (`POST \u002Fv2\u002Ffiles`, or the MCP upload tools) and use the returned URLs.\n\nNo actor photos? Design the cast with `varg-character-design` (archetypes, a model sheet per character on flat grey) and use each character's full-body crop as the profile card below.\n\n## 1. Consistency\n\n### 1.1 Character profile cards\n\nOne image-edit call per actor, from their photo:\n\n```tsx\nconst a1Card = Image({\n  prompt: {\n    text: \"Full-body character reference card of the exact same woman from the photo: same face, same grey hair, same body shape. \u003Cage\u002Fbuild>. She wears \u003CFIXED OUTFIT, described precisely>. Standing straight facing camera, neutral relaxed pose, arms at her sides. Plain light grey studio background, even soft lighting, head-to-toe framing, photorealistic\",\n    images: [A1_PHOTO_URL],\n  },\n  model: varg.imageModel(\"nano_banana_2_1\"),   \u002F\u002F edits when images are passed\n  aspectRatio: \"9:16\",\n});\n```\n\nRules that make cards work:\n\n- **\"exact same ... from the photo: same face, same hair, same body shape\"** anchors identity to the reference instead of the text description.\n- **Fix the outfit in the card.** The card's clothes become the outfit for the whole video. Pick it so the script works (e.g. arms or legs visible if scenes point at them). Never re-describe clothes differently later.\n- **Full body, neutral pose, plain studio background**: a clean card composites into any scene; a scene-styled card does not.\n\nThe pipeline was first built with Nano Banana Pro edit; `nano_banana_2_1` is its successor and varg's default. For the hardest identity cases try `gpt_image_2_5`.\n\n### 1.2 Location card\n\nOne text-to-image generation, **no people**:\n\n```tsx\nconst gymLoc = Image({\n  prompt: \"Empty modern gym interior with huge floor-to-ceiling windows... No people. ... photorealistic, 9:16 portrait framing\",\n  model: varg.imageModel(\"nano_banana_2_1\"),\n  aspectRatio: \"9:16\",\n});\n```\n\nAn empty location card lets every scene reuse the same walls, light and floor without characters from a previous scene leaking in.\n\n### 1.3 Scene composites\n\nEvery scene is one edit call with **all three refs**:\n\n```tsx\nconst SAME =\n  \"First reference image is Woman A (\u003Cshort recap of card>). Second reference image is Woman B (\u003Cshort recap>). Third reference image is the gym location with huge windows. Keep both women's faces, hair, bodies and outfits exactly as in their reference cards, and the gym exactly as in the location reference.\";\n\nconst scene2 = Image({\n  prompt: { text: `${SAME} \u003Cscene-specific framing, poses, gestures>`, images: [a1Card, a2Card, gymLoc] },\n  model: varg.imageModel(\"nano_banana_2_1\"),\n  aspectRatio: \"9:16\",\n});\n```\n\n- The `SAME` preamble is **shared verbatim** across all scenes. It names each reference by position (\"First reference image is...\") and pins identity.\n- Image edit models take several refs fine. For **Video**, one image = the first frame; several images switch to reference mode, which only some models support (see `varg-ai`).\n- Scene text only describes what changes: camera framing, poses, gestures, expressions.\n\n### 1.4 Cache discipline (critical)\n\nIdentical prompt + params = free cache hit, both locally and **server-side on api.varg.ai**. This is what makes the whole pipeline cheap:\n\n- Keep prompts **byte-identical** between the images-only script, the local template and the cloud template. Copy-paste, never retype.\n- The full 5-scene video re-rendered in the cloud for a few credits (the stitching fee) because every image, video and music prompt matched the earlier runs.\n- Editing one scene's prompt regenerates only that scene; everything else stays cached.\n\n### 1.5 Content-moderation gotcha\n\nBody-part wording such as \"rear\" or \"booty\" in an **image** prompt can silently stall generation for ~10 min (provider moderation). Rephrase to neutral anatomy: \"hips and thighs\". The same words inside spoken dialogue in **video** prompts passed fine.\n\nFitness and swimwear scenes trip filters on words, not pictures: \"girl\" instead of \"woman, 28\", \"tight\", \"bikini\" and short prompts without a garment description get blocked on GPT Image 2.5, FLUX 3 and Grok, while a catalog-style prompt (adult age as a number, each garment named, an activity, \"activewear brand catalog photo\") passed on all six image models we tried. A block can be random: retry once, then rephrase, then route. Test log, safe-phrasing table and routing: `varg-character-design` → `references\u002Fmoderation.md`.\n\n## 2. Photos-for-approval pass (images first)\n\nGenerate and show all stills **before** paying for any video. A still costs a few credits; a video clip costs ten times more or more. Approve the cheap layer first.\n\nStandalone script pattern (`\u003Cproject>-images.ts`, run with bun):\n\n```ts\nimport { Image } from \"vargai\u002Freact\";\nimport { createVarg } from \"vargai\u002Fai\";\n\nconst varg = createVarg({ apiKey: process.env.VARG_API_KEY! });\n\u002F\u002F ... define a1Card, a2Card, gymLoc, scene1..sceneN exactly as in 1.x ...\n\nasync function save(name: string, el: ReturnType\u003Ctypeof Image>) {\n  const resolved = await el;               \u002F\u002F await triggers generation (cached)\n  await Bun.write(`output\u002F\u003Cproject>\u002F${name}.png`, await resolved.file.data());\n}\n\n\u002F\u002F cards + location first, then scenes (scenes reference the cards)\nawait Promise.all([save(\"a1-card\", a1Card), save(\"a2-card\", a2Card), save(\"location\", gymLoc)]);\nawait Promise.all([save(\"scene1\", scene1), \u002F* ... *\u002F]);\n```\n\nKey facts:\n\n- `Image({...})` is a lazy thenable: `await` resolves it, `.file.data()` gives the bytes. Awaited results are cached, so the later video render reuses them for free.\n- Run it in the background (`nohup bun run \u003Cproject>-images.ts > images.log 2>&1 &`) and report the PID.\n- Show the user: the cards + location + every scene. Iterate on individual scene prompts (only the changed one regenerates) until approved.\n\nNo bun? Generate the same stills one by one with `POST \u002Fv2\u002Fimage` (or MCP `generate_image`), passing the references in `files`. Keep the prompts identical to the ones you will use in the render.\n\n## 3. Final render: cloud (`POST \u002Fv2\u002Frender`)\n\nAfter the stills are approved, wrap them in videos + composition and submit the whole TSX to the cloud. No bun or ffmpeg needed on the machine that submits.\n\n### 3.1 Video layer: image-to-video with native audio\n\nDialogue is written into the video prompt and the model voices it:\n\n```tsx\nconst vid2 = Video({\n  prompt: {\n    text: `Same gym with huge windows. \u003Cscene action, static camera>.\nThe trainer says matter-of-factly: \"\u003Cthe scene's line>\"\nSoft gym ambient sounds.`,\n    images: [scene2],                      \u002F\u002F exactly ONE image = the first frame\n  },\n  model: varg.videoModel(\"kling_v4\"),\n  audio: \"native\",                         \u002F\u002F generates and keeps the dialogue audio\n  duration: 4,                             \u002F\u002F integer seconds (kling_v4: 3-30)\n});\n```\n\n- Quote dialogue lines verbatim in the prompt; add a voice direction (\"clear encouraging voice\", \"matter-of-factly\").\n- Budget ~1 s of dialogue per ~2.5 words, then trim (3.3).\n- `kling_v3` is what this pipeline was first tested on and still works the same way (integer 3-15 s); `kling_v4` is the current Kling. `seedance_2_5` gives the best faces and lip movement at a higher price.\n- Native audio needs `audio: \"native\"`. A flat `providerOptions: { varg: { generate_audio: true } }` is silently dropped and you get a silent clip.\n\n### 3.2 Cloud TSX rules\n\n`POST https:\u002F\u002Fapi.varg.ai\u002Fv2\u002Frender` with `{\"code\": \"\u003Ctsx>\"}` (MCP: `render_video` with `code`):\n\n- **No imports, no `createVarg`**: `Render\u002FClip\u002FImage\u002FVideo\u002FMusic\u002FCaptions` and `varg` are injected as globals. Local LSP errors on this file are expected and fine.\n- Must `export default (\u003CRender ...>...)`; no named exports.\n- Same `varg.*Model()` ids as local.\n\nPrice it first (MCP `estimate_cost` with tool `render` and the code, or `POST \u002Fv2\u002Festimate` on each clip's request) and get the user's go-ahead. Then submit and poll:\n\n```bash\ncurl -s -X POST https:\u002F\u002Fapi.varg.ai\u002Fv2\u002Frender \\\n  -H \"Authorization: Bearer $VARG_API_KEY\" -H \"Content-Type: application\u002Fjson\" \\\n  -d \"{\\\"code\\\": $(jq -Rs . \u003C template-cloud.tsx)}\"\n# -> {\"id\":\"job_xxx\",...}; poll every 15s:\ncurl -s https:\u002F\u002Fapi.varg.ai\u002Fv2\u002Fjobs\u002FJOB_ID -H \"Authorization: Bearer $VARG_API_KEY\"\n# completed -> output.outputs[0].url  (s3.varg.ai mp4)\n```\n\nIf all prompts match a previous run, the cloud render costs a few credits (server cache) and finishes in ~2 min.\n\n### 3.3 Dynamic pacing: trim dead air\n\nGenerated clips have ~0.2-0.5 s of warm-up and often trailing silence. Detect speech boundaries on the generated clips, then cut tight:\n\n```bash\nffmpeg -i scene.mp4 -af \"silencedetect=noise=-35dB:d=0.35\" -f null -\n```\n\n```tsx\n\u003CClip cutFrom={0.2} cutTo={5.6} duration={5.4}>{vid1}\u003C\u002FClip>  \u002F\u002F cut right after the line ends\n```\n\nThis took the test video from 27 s to 22.4 s with no content loss.\nKeep `Music` `duration` an **integer**: a fractional total like 22.4 failed with a 422 on ElevenLabs music. Round down.\n\n### 3.4 Centered captions in the cloud (inline SRT)\n\n`\u003CCaptions>` transcribes `Speech` elements by itself. For dialogue that lives in a video's native audio, the tested route is two passes (newer vargai builds also accept `src={vid.audio}`; if the render rejects that with \"Captions src must be ...\", use this route):\n\n1. Render once without captions (cloud or local).\n2. Upload the audio and transcribe it on varg: `POST \u002Fv2\u002Ffiles` (upload) -> `POST \u002Fv2\u002Ftranscription` with `{\"model\":\"whisper_large_v3\",\"audio_url\":...}`. The result has word-level timestamps.\n3. Build a word-level SRT; fix Whisper mishears (product names, unusual words). Captions' `prompt` prop can also pass context hints to Whisper.\n4. Embed the SRT **inline** in the cloud TSX. The `srt` prop takes a file path, and template code runs on the render server, so write the file to `\u002Ftmp` at module eval:\n\n```tsx\nconst SRT_CONTENT = `1\n00:00:00,260 --> 00:00:00,320\nYou\n...`;\nconst SRT_PATH = `\u002Ftmp\u002Fproject-${Date.now()}.srt`;\nawait Bun.write(SRT_PATH, SRT_CONTENT);\n\u002F\u002F in \u003CRender>:\n\u003CCaptions srt={SRT_PATH} style=\"tiktok\" position=\"center\"\n          color=\"#ffffff\" activeColor=\"#a3e635\" wordsPerLine={3} fontSize={64} \u002F>\n```\n\nRe-submitting is nearly free (everything cached); only the composition re-runs.\nImportant: if you change any clip trim, **regenerate the SRT**, because the timings shift.\n\n## 4. QA: how to analyze what you generated\n\nAgents can't watch video, but they can read frames and measure audio.\n\n### 4.1 Visual QA: extract frames and READ them\n\nAfter every render (and every images pass), pull frames and inspect them as images. Never ship a video you haven't frame-checked.\n\n```bash\n# one frame per scene, at a timestamp INSIDE that scene (mind the trims)\nffmpeg -v error -ss 3  -i out.mp4 -frames:v 1 -vf scale=270:-1 check-s1.png -y\nffmpeg -v error -ss 9  -i out.mp4 -frames:v 1 -vf scale=270:-1 check-s3.png -y\nffmpeg -v error -ss 14 -i out.mp4 -frames:v 1 -vf scale=270:-1 check-s4.png -y\n```\n\n- Downscale (`scale=270:-1` .. `360:-1`): enough to judge, cheap on context.\n- Read several frames **in one parallel tool call**.\n- No local ffmpeg? The varg MCP `analyze_video` tool answers questions about a rendered video.\n\nVisual checklist per frame:\n\n- **Identity**: same face\u002Fhair\u002Fbody as the profile cards, in every scene.\n- **Outfit drift**: clothes must match the cards exactly (color, cut, shoes). Drift usually means a scene prompt re-described clothing: fix the wording, not the card.\n- **Location drift**: windows\u002Ffloor\u002Fprops consistent with the location card.\n- **Framing**: matches the script (full-body vs close-up); subjects not cropped at the wrong place; both actors present when required.\n- **Captions**: correct position (center), readable size, active-word color visible, no overlap with faces.\n- **Age\u002Fbuild drift**: edit models can age or slim a person between scenes. Compare against the card and regenerate the outlier scene only.\n\nAlso verify structure numerically:\n\n```bash\nffprobe -v error -show_entries format=duration,size -of default=noprint_wrappers=1 out.mp4\n```\n\nDuration must equal the sum of clip durations; a mismatch means a clip or music duration bug, not an encoding issue.\n\n### 4.2 Audio QA: silencedetect as ears\n\n`silencedetect` shows where speech actually lives in each generated clip:\n\n```bash\nffmpeg -i scene.mp4 -af \"silencedetect=noise=-35dB:d=0.35\" -f null - 2>&1 | grep silence_\n```\n\nRead it per clip:\n\n- silence at the **start** -> dead air \u002F late speech start -> raise `cutFrom`;\n- silence at the **end** -> line finished early -> lower `cutTo`;\n- **no silence** in an N-second clip -> speech fills it; if the line is long, the model may have rushed or truncated it: listen-check via transcription;\n- silence in the **middle** of a one-line clip -> possible stutter or split delivery.\n\nThen confirm the words: transcribe the final mix (3.4 step 2) and diff the transcript against the script. This catches mispronunciations by the video model and mishears by Whisper. Fix mishears in the SRT; regenerate the clip only if the model actually said the wrong thing.\n\n### 4.3 When to regenerate vs when to trim\n\n| Symptom | Fix |\n|---|---|\n| Dead air, late start, early finish | Trim with `cutFrom`\u002F`cutTo`: free |\n| Wrong\u002Fgarbled spoken line | Regenerate that Video |\n| Identity\u002Foutfit drift in one scene | Regenerate that scene Image, then its Video |\n| Caption timing off after re-trim | Regenerate the SRT only (one transcription) |\n\n## 5. No script? Analyze the reference video first\n\nWhen the user sends a reference video instead of a script, reverse-engineer the same A\u002FL\u002FScene structure this skill consumes. Do this **before** touching any generation. (`media-download` fetches a reference by link.)\n\n### 5.1 Extract structure\n\n```bash\nffprobe -v error -show_entries format=duration -of default=noprint_wrappers=1 ref.mp4\n\n# scene cut candidates (hard cuts)\nffmpeg -i ref.mp4 -vf \"select='gt(scene,0.35)',showinfo\" -f null - 2>&1 | grep showinfo | grep pts_time\n\n# speech map\nffmpeg -i ref.mp4 -af \"silencedetect=noise=-35dB:d=0.35\" -f null - 2>&1 | grep silence_\n```\n\nScene-cut timestamps + speech gaps together give the beat structure: how many scenes, how long each, where dialogue sits.\n\n### 5.2 Sample frames per scene\n\nExtract 1-2 frames inside each detected scene and read them (parallel call):\n\n```bash\nffmpeg -v error -ss \u003Ct> -i ref.mp4 -frames:v 1 -vf scale=360:-1 ref-s1.png -y\n```\n\nFrom the frames, document for each scene:\n\n- **Actors**: who is on screen, rough age\u002Fbuild\u002Fhair -> becomes A1, A2...\n- **Wardrobe**: what each actor wears (this defines the profile-card outfits).\n- **Location**: environment, light, key props -> becomes L1 (L2...).\n- **Framing**: full-body \u002F medium-wide \u002F close-up; front or back view; camera static or moving; who is closer to camera.\n- **Action + gesture**: what each actor does (squats, points at X, thumbs-up).\n\n### 5.3 Get the dialogue\n\nTranscribe the reference audio (same varg call as 3.4). Map each line to its scene via the timestamps. If there's no speech, note the music mood and on-screen text instead.\n\n### 5.4 Write the script, then confirm\n\nAssemble the standard format and **show it to the user for approval** before generating anything:\n\n```\nActors:\n- A1: \u003Cdescription from frames>  - C1: \u003Cwardrobe>\n- A2: \u003Cdescription>\nLocation:\n- L1: \u003Cdescription>\nScene 1 (Ns): \u003Cframing>. \u003Caction>. A2 says: \"\u003Cline>\". L1\nScene 2 ...\n```\n\nFrom here the pipeline is identical: cards (1.1) -> location (1.2) -> composites (1.3) -> approval pass (2) -> videos + cloud render (3). Use the reference's measured scene durations as the initial `duration` values, and its framing notes verbatim in scene prompts.\n\n## Cost reference (5 scenes, ~22 s at 9:16)\n\nRough credits from the varg catalog (October 2026). Always quote the user the exact figure from `POST \u002Fv2\u002Festimate` \u002F `estimate_cost`, in credits.\n\n| Step | Credits (~) |\n|---|---|\n| 8 images (2 cards + location + 5 scenes), `nano_banana_2_1` | ~65 |\n| 5 Kling clips with native audio, 4-6 s at 720p | ~400-500 |\n| Music bed (`mureka_9_5`) + one transcription | ~30 |\n| Cloud re-render with everything cached | a few (stitching fee) |\n| Iterating one scene image \u002F one scene video | ~8 \u002F ~80-100 |\n\n## Related skills\n\n`varg-ai` (sign-in, models, render formats), `varg-character-keynote` (one speaker on a stage), `varg-podcast-video` (two seated characters in dialogue), `varg-previz` (plan many shots and camera moves first), `media-download` (fetch a reference video), `video-skills` (find a skill for another format).\n",{"name":5,"license":15,"metadata":52,"description":7,"allowed-tools":61,"compatibility":62},{"author":53,"version":54,"openclaw":55},"vargHQ","1.0.0",{"homepage":56,"requires":57},"https:\u002F\u002Fvarg.ai",{"anyBins":58},[59,60],"curl","bun","Bash(bun:*) Bash(bunx:*) Bash(curl:*) Bash(jq:*) Bash(ffmpeg:*) Bash(ffprobe:*) Bash(nohup:*) Read Write Edit","Needs a VARG_API_KEY (see the varg-ai skill) or the varg MCP connection (https:\u002F\u002Fmcp.varg.ai\u002Fmcp). The images-only approval script needs bun; the final render runs in the cloud with curl. QA uses ffmpeg\u002Fffprobe locally.","main","2b63f856ce19c2a797929cacabd9e0fc936812fa","plugins\u002FvargHQ\u002Fskills\u002Fvarg-consistent-characters",491,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins","https:\u002F\u002Fhol.org\u002Fplugins\u002Fbest-codex-plugins","https:\u002F\u002Fgithub.com\u002Fhashgraph-online","https:\u002F\u002Fhol.org","Open-source standards and developer tools for interoperable AI agents: discovery, identity, communication, payments, privacy and security.","2026-11-14 14:06:13",{"source":8,"skillName":5,"commands":74,"prompt":84,"zipAvailable":85},[75,78,81],{"tool":76,"command":77},"npx","npx skills add hashgraph-online\u002Fawesome-codex-plugins --skill varg-consistent-characters",{"tool":79,"command":80},"bunx","bunx skills add hashgraph-online\u002Fawesome-codex-plugins --skill varg-consistent-characters",{"tool":82,"command":83},"pnpm","pnpm dlx skills add hashgraph-online\u002Fawesome-codex-plugins --skill varg-consistent-characters","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002FvargHQ\u002Fskills\u002Fvarg-consistent-characters\u002FSKILL.md",true,[],[88,97,105,113,121,129],{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":92,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":96},"01M7DV332YQMGD8WTRJ9DZQXTB","worktable-review-with-annotations","Create, inspect, reply to, update, or resolve situated Worktable annotations. Use when feedback belongs to a Doc section, block, text range, or HTML Doc, or when the user asks to address existing review feedback and coordinate a requested artifact edit.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-review-with-annotations\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-review-with-annotations\u002FSKILL.md",[],[],"2026-11-15 14:19:17",{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":101,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":96},"01M7DV2YGHCZVBEB1AHYB6CGHE","worktable-find-and-synthesize-context","Find, assess, and synthesize context from Worktable Docs and Records. Use when the user asks what a workspace contains, wants workspace-backed research, needs current facts reconciled across artifacts, or requests a cited synthesis without changing workspace state.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-find-and-synthesize-context\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-find-and-synthesize-context\u002FSKILL.md",[],[],{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":109,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":96},"01M7DV2TQN1H11ME2SAF9MXN7V","worktable-create-or-update-docs","Create, edit, rename, or maintain narrative Worktable Docs. Use for plans, decisions, briefs, notes, research syntheses, and other durable prose when the user requests a new artifact or a safe update to an existing one.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-update-docs\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-update-docs\u002FSKILL.md",[],[],{"skillNo":114,"name":115,"nameZh":6,"description":116,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":117,"sourceHtmlUrl":118,"categories":119,"keywords":120,"publishedAt":96},"01M7DV2NE452NEA8TXATKSNAJH","worktable-create-or-manage-records","Organize repeated structured information as durable Worktable Records. Use for research datasets, inventories, directories, pipelines, trackers, and other collections whose items need identity, fields, filtering, links, validation, or individual updates.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-records\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-records\u002FSKILL.md",[],[],{"skillNo":122,"name":123,"nameZh":6,"description":124,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":125,"sourceHtmlUrl":126,"categories":127,"keywords":128,"publishedAt":96},"01M7DV2HENRBR2KSCVM064WH0Y","worktable-create-or-manage-interactive-html","Create or improve durable visual and interactive artifacts in Worktable. Use for dashboards, trackers, calculators, mini apps, slide presentations, and other requests that benefit from a polished HTML Doc.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-interactive-html\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-interactive-html\u002FSKILL.md",[],[],{"skillNo":130,"name":131,"nameZh":6,"description":132,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":133,"sourceHtmlUrl":134,"categories":135,"keywords":136,"publishedAt":96},"01M7DV2CADR5R7RQFFYPNAGR9Y","worktable-collaborate-threads","Read, answer, and hand off messages through durable Worktable threads with users or other authenticated participants. Use when the user asks to check addressed messages, reply to a participant, continue a thread, request another agent's help, wait briefly for a response, or watch for comments and messages sent to the agent.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-collaborate-threads\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-collaborate-threads\u002FSKILL.md",[],[]]