beauty-generation-api

**Judgement:** A specialized, well-documented portrait-generation skill that is strong for creative and professional-style portraits, but relies entirely on an external third-party API and operates in a sensitive domain (attractive human faces), so it requires cautious, policy-aware use. **Key strengths:** - Focused, high-level control over human portraits (styles, nationalities, moods, backgrounds, clothing) and fast generation via an HTTP API. - Good developer ergonomics with clear curl examples, a robust Python script template, and explicit guidance on quota checking and polling. - Explicit content-safety rules (no minors, nudity/sexual content, violence, hate, illegal activity, or undisclosed deepfakes) and emphasis on 18+ only subjects. **Main risks / limitations:** - Depends on an external, free service (gen1.diversityfaces.org) with its own uptime, rate limits, and policy enforcement; failures or changes are outside the agent’s control. - Domain risk around beauty and attractiveness (potential for biased outputs, unrealistic beauty standards, and user attempts at borderline sexual or identity-based content). - Users must handle their own API keys; mishandling keys or sending sensitive personal data to the third-party endpoint would be a privacy/security concern. - Early-development status of the skill and API implies potential instability or breaking changes. **Recommended scenarios:** - Creative character design, avatar creation, and fashion/portrait concepting where subjects are clearly fictional or non-identifying. - Rapid prototyping of UI/UX concepts, marketing mockups, or illustrative assets that need diverse, professional-looking portraits without tying to real individuals. - Workflows where technically inclined users are comfortable managing API keys, quotas, and optionally writing local scripts to automate generation. **Use with caution when:** - Requests approach sensitive areas (e.g., realism close to specific real people, sexualization, or demographic targeting); strict adherence to the stated content rules is required. - Applications might be mistaken for real, identifying photography or deepfakes; transparency and disclaimers are important, and some use cases should be declined entirely.