Edit a video in chat and export the finished file

A developer made an OpenClaw skill that turns chat requests into backend video edits and returns the rendered file.

Done with OpenClaw· Advanced

What they did
They built an OpenClaw skill with a SKILL.md file that tells the agent how to handle video-editing requests. Simple actions like upload, credit checks, and export go to API calls, while other edits stream to the NemoVideo backend over SSE. The agent also removes the backend’s GUI-style directions, then compares timeline state before and after if an edit returns no text.
How it went
In the example, asking for Chinese subtitles and an mp4 export finished in about 90 seconds and produced an mp4 with burned-in subtitles. Around 30% of edits return no text, and changing aspect ratio requires starting over with a full regeneration. Retry silent edits caused duplicate changes and double credit charges.
Tips
Have the agent show the planned edit before it makes irreversible changes. Avoid retrying a request that comes back silent, and remember that triggering export from the backend starts a fresh generation that uses credits even though export itself is free.

Try it yourself with OpenClaw

I’ll upload a video and ask for edits in chat. First, tell me exactly what changes you plan to make to [describe the edit, such as add subtitles in Chinese and export as mp4], and wait for my approval before you do anything that could overwrite, submit, send, delete, or charge credits. Then carry out the approved edit through [your video editor or backend], keep my original file unchanged, and return the finished export once it’s ready.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on dev.to ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.