Pick Halloween Costume · Open Skill 公開技能
Pick Halloween Costume
Pick Halloween Costume 是 Digen 上的公開 Open Skill。開啟後即可了解此工作流能製作什麼。
Roleplay
0 0 0
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4284957/2026-10-16/j7y7jfu3.gif
# Pick Halloween Costume Pick your monster. Your photo turns into a Halloween costume transformation clip — zombie, vampire, demon or your pick — with a clear before-to-after morph and spooky atmosphere. ## When to use - The user wants to become this look in a video: a short Halloween costume transformation video from their photo with a chosen monster style - A variant on a previous result (another style, length or orientation) Not for: Still images only, or animating a picture with no transformation. ## Inputs and gates | Input | Required | Default | Gate | |---|---|---|---| | Photo | Yes | — | One adult with the face clearly visible. Ask for another photo if the face is tiny, blurred, heavily covered or in profile | | Costume | No | `zombie` | zombie (default), vampire, demon, or user-described | | Duration | No | `6` s | 6 or 10 | | Orientation | No | `9:16` portrait | Other ratios on request | - If the person may be under 18, stop and decline politely. Real-person photos must be of adults the user has the right to use. - Video does not allow NSFW content: keep wardrobe fully covering and refuse sexual requests. Suggestive fashion looks stay as still images. ## Steps 1. Read `references/transformation-reveal.md` before writing prompts. 2. **Identity read**: `recognize_image(image_url=<photo>, prompt='Describe this adult for identity preservation: face shape, eye shape and color, brows, nose, lips, skin tone, apparent age range, hair color and length, distinctive marks, glasses. One compact line.')`. 3. **Before frame**: `generate_image(prompt=<BEFORE>, input_urls=[photo], model='i2i.hd', orientation='portrait', aspect_ratio='9:16', name='Pick Halloween Costume before', entity_type='work_in_progress')`: the person in their own clothes, already in the final scene, so both frames share the set. 4. **After frame**: `generate_image(prompt=<AFTER>, input_urls=[before frame, photo], model='i2i.hd', orientation='portrait', aspect_ratio='9:16', name='Pick Halloween Costume after', entity_type='work_in_progress')`: the hero key art. 5. **Self-check, one retry per frame**: if the face no longer reads as the same person, regenerate with the identity clause first and fewer face effects; if the after frame looks like mild makeup, regenerate with a bolder silhouette and stronger costume change. 6. **Reveal video**: `generate_video(prompt=<MOTION>, input_urls=[before frame, after frame], model='fl2v', orientation='portrait', aspect_ratio='9:16', duration=<d>, name='Pick Halloween Costume', entity_type='final_videos')`. If `fl2v` fails, fall back to `generate_video(prompt=<MOTION_I2V>, input_urls=[before frame], model='i2v', ...)`. 7. Reply in one or two sentences, e.g. "Done! Here's your pick halloween costume. I can also try vampire if you like." Results appear on the canvas; never paste URLs. ## Prompt templates **BEFORE** ```text The same adult from the reference: <identity line>. Keep their real face, hair and everyday outfit exactly. They stand in the final scene of this concept: the same adult transforms into a full Halloween <costume: zombie (default), vampire, demon, or user-described> look: complete costume, makeup and themed silhouette while identity stays locked; dramatic spooky lighting, fog or candle accents, clear costume reveal motion from everyday clothes into the finished look, playful not gory, fully clothed, but before any change, wearing their own clothes, calm expression toward camera. Medium-full shot, eye-level, cinematic photoreal still, no text or lettering anywhere. ``` **AFTER** ```text The same adult from the references: <identity line>. Same face structure, eyes, skin tone and recognizable likeness; no beautification or face reshaping. Intentional changes: the full look of this concept: the same adult transforms into a full Halloween <costume: zombie (default), vampire, demon, or user-described> look: complete costume, makeup and themed silhouette while identity stays locked; dramatic spooky light