software-tutorial-video-maker

Provide your screenshots and rough notes, and the Skill writes a tutorial brief and storyboard, generates one narration clip per segment with the speak tool, then assembles the final video with ffmpeg using audio-led timing.

The Skill uses ffmpeg for image-to-video segment creation, audio muxing, optional light transitions, segment concatenation, and final export normalization. The exact ffmpeg commands used are recorded in video-package.md for reproducibility.

It needs the source screenshots or frames plus rough instructional notes. If screenshots are missing, out of order, or the notes are too vague to support truthful narration, the Skill stops and asks for clarification rather than inventing content.

Yes. The speak tool accepts a language parameter when a specific narration language is requested, and one consistent voice is used across segments unless the tutorial explicitly requires variation.

The Skill only narrates what is visible in the supplied screenshots and will not invent unseen menus, cursor paths, or success states. It defaults to still-image assembly with cuts or light fades, adding no fake cursor movement or cinematic motion unless requested.