| name | orchestrating-restoreassist-video-series |
| description | Use when producing, scripting, or shipping the RestoreAssist onboarding + walkthrough video SERIES as a batch (not one clip) — the setup-wizard / getting-started tutorial / onboarding-welcome videos need scripts written, narrated, rendered, captioned, and wired into the live app for distribution; when picking up where a prior session left off across many slugs; or when a polished walkthrough must land on YouTube and resolve in <VideoExplainer>. |
Orchestrating the RestoreAssist Video Series
Overview
This is the conductor over the whole onboarding + walkthrough video series. Each individual video is produced with the per-video pipeline; this skill sequences the set — tracking every slug's state, driving the next ones to "done", and landing them in the live app on YouTube unlisted embeds for distribution.
Core principle: the series advances one status-derived wave at a time. You never start from zero — you read the repo to see what each video already has, then do only the missing steps. A video is "done" only when its slug plays in the running app, not when a file exists.
REQUIRED SUB-SKILL: Use producing-restoreassist-videos for the mechanics of any single video (the five artifacts, ElevenLabs voice, render-all wiring, caption gotchas). This skill does NOT repeat that — it orchestrates many runs of it.
The series (this is "all these videos")
Defined by registry slug → Remotion composition id → file stem. Stems are the MP3/MP4/VTT/public_id filename; slugs key the registry and analytics. Captions are the trap: getCaptionUrl(slug) looks up CAPTION_REGISTRY[slug] by the slug, but the file on disk is stem-named — and the registry is currently keyed inconsistently (onboarding by slug, wizards by stem). See the caption-key note below.
