Dialogue time share titles
Measure each speaker's voiced time from labeled cue intervals, merging overlaps within a speaker before displaying the results.
Preview
Rendered with version 1.0.0
What you provide
- Speaker-labeled start and stop times in seconds for one transcript.
What you get
- A 960 × 540 H.264 MP4 clip and a JSON report with the verified result.
Requirements
- Python 3.9+, FFmpeg with libx264, and librsvg available locally.
- No provider account required.
Installation
- For Codex, extract the skill directory into .agents/skills in your project, then reopen your agent session.
- Extract the ZIP into a new folder and open the skill directory.
- Read SKILL.md and check the required local programs.
- Run python3 render.py examples/default.json preview.mp4.
- Copy an example JSON project, replace its input values and render to a new filename.
License and release notes
Local edits and finished client videos permitted. Redistributing the package or handing off its editable sample assets requires permission.
1.0.0 · Initial release with two verified editable projects.