Automation Agent: editable rough cuts in Premiere Pro

What the rough-cut recipe produces

The documented recipe takes transcribed clips, an editing brief and a target duration. It requires the Python Runtime Pack and a configured Workflow AI provider. After input review and validation, it creates a new sequence with selected segments and markers for inspection.

Spoken-content rough cut recipe ↗

Know which inputs leave the computer

The rough-cut workflow sends the brief and transcripts to the configured provider after review. The maker describes face pixelation as a separate local workflow that leaves editable clips and effects. Local face processing does not make every workflow local.

Spoken-content rough cut recipe ↗Automation Agent for Premiere Pro ↗AI Assistant for Premiere Pro: Reusable Workflows ↗

Usability still needs a trial

Two visible comments under the maker’s demonstration give opposing accounts: a beta tester describes a reusable social-video workflow, while another editor reports difficulty completing a task. These are individual reports, and the editing workflow was not executed here.

AI Assistant for Premiere Pro: Reusable Workflows ↗

What you can do

1. Read the rough-cut recipe and use the trial on a short, non-sensitive project whose desired selections you can check by hand. Open reference ↗

Where the evidence stops

The documentation and video metadata were checked. Cut quality, setup effort and Windows-on-ARM behavior remain unverified; the listing says that platform is untested.

Sources read. Public sources read October 23, 2026 UTC. No external product workflow, model inference or reported experiment was executed.

Sources & further reading