Your Very Own Otome Game Male Lead · publiczny skill Open Skill
Your Very Own Otome Game Male Lead
Your Very Own Otome Game Male Lead to publiczny Open Skill w Digen. Otwórz go, by zobaczyć, co tworzy ten workflow.
Manhua
60 7 18
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/3947739/2026-08-26/jaquasq4.png
# Adult First-Person Flirty Chat Video Maker (18+) ## Hard Requirements (read first) - All "must / forbidden / never / highest-priority" rules are centralized in the reference file `references/hard-requirements.md` - Before starting, you MUST call read_skill_file(name='adult-chat-video-maker_2', path='references/hard-requirements.md'), read it, and strictly follow every rule - If body steps conflict with the hard requirements, the hard requirements win - **Highest priority**: This skill is strictly for 18+ adult characters and adult topics. Strictly forbidden to generate any image that looks like a minor or any dialogue/content involving minors. If the user input implies an underage age, refuse immediately and remind that it must be 18+. ## When to use - The user uploads a photo of a person and wants to create an 18+ adult character (age specified by the user, default 22-28, male or female), and make a first-person-POV flirty/romantic/erotic chat vertical short video - Supports various adult explicitness levels (light flirting, moderate erotic, heavy sensual, etc.), explicitly specified by the user - Also use when the user only wants to redo one stage (character reference image, chat lines, scene storyboards, one line's shot, BGM, final mix) ## Self-check and confirmation after generation (do this at every generation stage) - After each image, self-check with recognize_image - After each video, self-check with recognize_video - Self-check focus: clearly adult (no minor features); character consistency; art style consistency; frame contains only the character; storyboard composition/lighting/art style completely consistent; character–camera interaction natural (looks into the lens, approaches, touches, etc.) and the character stays at the same location/position while interacting; dialogue box correctly shows character name + lines (user has no lines, no user dialogue box); chat topic follows the user's initiated topic and the character talks to the user conversationally with natural pauses; line explicitness meets the user's requirements; voice consistent throughout and matching gender; lip sync; fixed camera; natural transitions - You MUST honestly and completely report all problems found in self-check (including any age risk) to the user, then via set_guidance provide "Regenerate" and "Continue to next step" buttons, and let the user decide ## Steps ### 1. Collect input and confirm requirements - Ask the user to upload 1 photo of a person - **Mandatory age confirmation**: must be clearly ≥18 (default 24). If the user says under 18 or uses any description implying underage, refuse immediately and say "This skill only supports adults 18+, please provide an adult image" - Confirm character gender, personality, specific explicitness level (light flirty / moderate erotic / heavy sensual / specific play, etc.), scene, what the character calls the user, voice preference (must match gender, can be an adult voice) - Use recognize_image to analyze the photo, make sure the appearance is clearly adult, and write the features into every prompt - Strictly follow the age restrictions in the hard requirements ### 2. Generate character reference image - generate_image, model='i2i.hd.lite', input_urls=[original photo], orientation='portrait' - prompt essentials: 18+ mature adult look (strengthen adult facial contours, eyes, body curves, aura), fully consistent with the original photo's facial features/face shape/hairstyle/clothing/art style, no childlike or underage features at all; expression based on the user-specified personality and explicitness level (may include seductive, sensual adult expressions); natural standing, composition above the calf, clean background - name='Character reference image', entity_type='characters' - After generation, immediately self-check with recognize_image (clearly adult? looks like the person? aura matches the explicitness level?) and report the result fully to the user - Then ask whether the user is s