scrollclaw-assemble

Assemble UGC videos by stitching A-roll first, unifying the voice track, inserting B-roll as visual-only segments, applying post-production realism, and overlaying resolution-matched TikTok-native caption PNGs last to preserve realism.

Correct stage ordering stitches A-roll first, orchestrates the continuous voice track, inserts B-roll as visual-only segments, applies post-production realism, and renders TikTok-native captions last to maintain cinematic consistency.

You can run stitch-only, post-only, captions-only, or full-assemble workflows independently. Mode-based execution lets you isolate specific pipeline stages like TikTok-native caption rendering or post-production realism without re-stitching.

Normalize resolution and frame rates by enforcing cross-engine color and lighting matching during post-production. This validates Sora A-roll and Kling B-roll hybrid formats against realism checklists to catch AI tells early.

Caption PNG overlays appear misaligned if they are not generated at the exact video resolution. TikTok-native caption rendering requires resolution-matched transparent PNGs using TikTok Sans presets and safe-zone compliance to align correctly.

The assembly pipeline supports both single-clip and multi-clip talking-head formats, including hybrid formats that combine Sora A-roll with Kling B-roll, stitching them with continuous voice tracks and post-production realism.