Source and place accurate b-roll cutaways against spoken beats in talking-head videos.
npx skills add https://github.com/louisedesadeleer/b-roll-finder --skill find-broll
Please help me install this Agent Skill. Skill: find-broll Source: https://github.com/louisedesadeleer/b-roll-finder/tree/main Command: npx skills add https://github.com/louisedesadeleer/b-roll-finder --skill find-broll
This Skill removes the guesswork from sourcing b-roll for talking-head videos by connecting each spoken beat to accurate, relevant, and on-brand visual footage.
Use the find-broll skill with a transcript or video file to propose an annotated b-roll plan, source approved candidates, and place the selected cutaways on the exact spoken words.
High-intent search queries and answers about installing and using this skill.
To source b-roll for talking-head videos, this Skill analyzes word-level transcripts to understand context, scores relevant visual candidates, and anchors cutaways precisely on the spoken beats. It routes searches to authoritative entities, receipts, and product footage.
Yes, you can place b-roll cutaways using podcast video files or transcripts. The Skill processes word-level transcriptions to identify thesis points, sources matching footage, and renders full-bleed cutaways anchored just after the relevant spoken word.
Automated b-roll placement requires word-level transcription to achieve precise timing. The Skill relies on this granular transcript data to interpret the video's thesis, identify what each line is about, and anchor cutaways on the exact spoken words.
The best way to match b-roll footage in explainers is scoped media sourcing. The Skill interprets the transcript, routes moments to authoritative entities or concept graphics, scores candidates, strips audio, and validates the rendered frames.
To render and validate b-roll cutaways locally, you need ffmpeg, ffprobe, Pillow, websockets, and Google Chrome. These dependencies support silent full-bleed rendering, precise timing, and visual render validation.
B-roll footage often mismatches context when relying on generic keyword-matched stock. This Skill prevents this by understanding the video's thesis, interpreting each line, and sourcing contextually relevant guest clips, product footage, and supporting receipts.
barefootfordTranscribes video audio into timestamped JSON transcripts with word-level timing using WhisperX.
het8802Automate B-roll search, FFmpeg video editing, and asset ledger tracking.
gooseworks-aiGenerate full-bleed motion video ads from briefs and brand assets using Hyperframes and FFmpeg.
spindle79Plans multi-clip Seedance video sequences with continuity tracking and audio-matched segments.
