GlamMode Studio · सार्वजनिक Open Skill
GlamMode Studio
GlamMode Studio Digen पर एक सार्वजनिक Open Skill है। खोलें और जानें कि यह वर्कफ़्लो क्या बनाता है।
Anime
13 5 2
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4247767/2026-10-10/xpht34qy.jpg
# GlamMode Studio Create a photorealistic vertical (9:16) fashion / swimwear campaign video with a consistent on-screen subject, from a brief or a character description. ## Reference files - `references/example-clips.md`: Ready-made example clip briefs (subject, outfit, setting, action) with ready-to-use image and video prompts. Read it when the user wants to pick a preset example clip style or needs inspiration. ## Preset example clips Three playable example clips are bound as preset assets — refer to them by key when the user wants a preset look or wants to see/redo an existing example: - `example_pool_walk` — Poolside Walk: elegant woman in a white one-piece by a luxury resort pool. - `example_beach_golden` — Beach Golden Hour: woman in a tropical print bikini and straw hat on a golden-hour beach. - `example_yacht_deck` — Yacht Deck: relaxed woman in an off-shoulder swimsuit on a yacht over turquoise water. These are reference videos for style/mood; to produce a new clip, still generate a fresh character reference image and run the video step (see Steps). ## When to use - The user wants a photorealistic vertical fashion / swimwear campaign video (model walking beside a pool, adjusting sunglasses, smiling at camera, etc.). - The user wants to redo/refine an existing clip or change one detail (outfit, motion, setting, duration). - The user wants to use a preset example clip style or start from an inspiration idea. ## Steps 1. For a multi-step run, first call `write_todos` to list the steps and update it as you progress (one `in_progress` at a time). 2. Clarify the brief if not fully specified: subject (a fictional adult woman, 25+, photorealistic), outfit, setting, key motion/action, and duration (default 12s, accept 10–15s). If the user wants a preset style or has no idea, offer one of the preset example clips (see `Preset example clips`) or read `references/example-clips.md`. Keep content tasteful — video generation does not support NSFW/adult content. 3. Generate a character reference image with `generate_image`: - `model='t2i'`, `orientation='portrait'`, `resolution='1080P'` - `name='Character Reference'`, `entity_type='characters'` - Prompt: photorealistic shot of the subject in the specified outfit and setting, consistent appearance, cinematic color grading, no text or watermark. This image locks the subject's consistent look for the video. 4. Generate the video with `generate_video`: - `model='ref2v'`, `input_urls=[the reference image asset_id]`, `orientation='portrait'`, `duration=12` (or the user's requested 10–15s), `resolution='1080P'` - `name='Fashion Video'`, `entity_type='final_videos'` - Prompt: describe the motion — e.g. "She walks confidently beside the pool, adjusts her sunglasses, and smiles toward the camera. Gentle camera tracking, warm sunlight, realistic fabric movement, cinematic color grading, polished swimwear campaign aesthetic, consistent appearance, no text or watermark." 5. Show the result and call `set_guidance` to suggest next actions (e.g. "Change the outfit", "Different setting", "Longer/shorter duration", "Make a new clip"). Offer to redo/refine: change the outfit, motion, setting, or duration, or regenerate the reference image for a different look. ## Notes - Video generation does not support NSFW/adult content on any channel; keep subjects fictional adults and content tasteful/swimwear-appropriate. - The intermediate character reference image will be visible on the canvas alongside the final video — that's expected, not an error. - Report failures honestly and never fabricate URLs; if a generation fails, retry once with a clarified prompt, then tell the user what happened. - Keep prompts in English and specify "no text or watermark" to avoid burned-in captions. - When the user changes one detail on a finished clip, regenerate the reference image only if the subject's look changes; otherwise reuse the existing `Character Reference` image and only adjust the motion/se