Video Repurpose Pipeline

@shaunak/video-repurpose-pipeline

Turn long-form videos into semantically complete, captioned vertical clips with adaptive framing and deterministic QA.

An agent-orchestrated, script-first workflow for video ingestion, word-timed transcription, semantic clip selection, exact boundary validation, adaptive face-aware vertical framing, phrase-timed captions, FFmpeg rendering, deterministic QA, and Content Lab delivery.

Read SKILL.md instructions · Full detail JSON

Files

SKILL.md
config/default.json
references/INSTALL.md
references/platform-presets.md
references/SPECIFICATION.md
references/youtube-ingestion.md
schemas/candidates.schema.json
schemas/edit-manifest.schema.json
schemas/publish-request.schema.json
schemas/qa-report.schema.json
schemas/source-analysis.schema.json
schemas/story-units.schema.json
schemas/transcript.schema.json
scripts/analyze_source.py
scripts/build_edit_manifest.py
scripts/common.py
scripts/detect_visuals.py
scripts/doctor.py
scripts/ingest_source.py
scripts/normalize_transcript.py
scripts/publish_content_lab.py
scripts/qa_outputs.py
scripts/rank_complete_clips.py
scripts/render_variants.py
scripts/run_pipeline.py
scripts/segment_story_units.py
scripts/setup_toolchain.py
scripts/transcribe.py
scripts/validate_boundaries.py

Discover · Public registry guide