What problem does it solve? Producing narrated audio and video content — guided meditations, sleep essays, or narrated documentaries — normally requires paid TTS APIs, a GPU, or manual audio editing. This Skill runs the entire pipeline locally with open-source tools, from writing the script to delivering a finished MP4. ## Core Features & Use Cases - Three content modes: guided meditations with deliberate silence pacing, long-form sleep essays with continuous slow narration, and awake casual-essay documentaries with many licence-verified images and original diagrams. - Fully offline pipeline: Kokoro ONNX text-to-speech plus ffmpeg rendering, with no API keys, no GPU, and resumable segment generation. - Quality gates built in: an enforced de-AI/fact-check convergence loop blocks audio generation until essay prose passes review, and delivery includes YouTube description, chapters, and tags. - Use Case: Ask for a 20-minute narrated video essay on an obscure historical topic; the Skill researches and writes the essay, verifies it, narrates it at the right pace, sources licensed images, renders the slideshow, and delivers the MP4 with publishing metadata. ## Quick Start Create a 5-minute guided breathing meditation video with a calm female voice over a misty lake backdrop.
