Back to catalog / Agent workflows

Turn video content into a narrated summary clip

Extract representative video frames, create a description and narration script, then synthesize a voice track.

Goal

Extract representative video frames, create a description and narration script, then synthesize a voice track.

Use OpenCV to sample frames and verify coverage of key scenes. [source-openai-create-narrated-video-summary]
Draft a concise narration, synthesize with TTS, and align audio to timestamps. [source-openai-create-narrated-video-summary]
Check facts, pacing, and rights; do not imitate a real person’s voice. [source-openai-create-narrated-video-summary]

Check that selected frames cover the beginning, middle, and end, narration fits the clip, and the exported audio plays.

If a key scene is missing, add representative frames and regenerate the description; for audio drift, inspect sample rate and segment timestamps.

Python, OpenCV, OpenAI vision and TTS APIs; process licensed video and check voice/likeness rights.

Uploads approved video or sampled frames to vision/audio APIs and calls speech synthesis, which may incur charges; writes output only to the designated local directory.

Run the source on isolated test data, then check outputs and dependency versions before adapting it.

These values describe the upstream repository, not popularity or likes for an individual skill or playbook.

No metrics snapshot is available yet.

License: MIT