What problem does it solve? Turning long tutorial videos, lessons, or screen recordings into readable study notes normally requires manual watching, pausing, screenshotting, and writing. This Skill automates that entire pipeline by understanding the video's speech, actions, and on-screen text, then producing a searchable, illustrated PDF note. ## Core Features & Use Cases - Omni-based video understanding: Analyzes actions, speech, visible text, and timestamps across chunked long videos using a single Omni model. - Illustrated PDF generation: Extracts screenshots near step timestamps locally and renders a searchable PDF note with visual evidence. - Configurable quality and budget: quality_profile controls chunk duration and sampling, while time_budget_seconds bounds model request time. - Graceful degradation: Falls back to a simpler document from partial evidence, omits failed screenshots, and returns structured failures instead of fabricated notes. - Use Case: A student records a 40-minute software tutorial and asks the agent to convert it into a PDF note; the tool transcribes key steps, captures screenshots at each step, and outputs a single illustrated document. ## Quick Start Use the omni video2note skill to convert my local file /videos/python-tutorial.mp4 into an illustrated PDF note saved at /notes/python-tutorial.pdf.
