Skill UI

This advanced skill generates high-quality, natural speech from text entirely offline using the VoxCPM2 model. Supporting 30 languages, it offers three modes: standard text-to-speech, descriptive voice design, and advanced voice cloning from reference audio. All processing runs on the local device (optimized for Apple Silicon), ensuring zero API calls, zero cost, and maximum privacy. Ideal for creating professional voiceovers, podcasts, and video narration without internet dependency.

Verify citations and references in documents against real sources using the Stipple API. Checks resolution, arithmetic, and unsupported claims. Use before publishing research reports, whitepapers, or academic writing.