Tutorials

Use AI Agents from Your CLI to Create Product Videos While You Code

Let AI agents handle your product video creation from the command line. Record, edit, brand, and export - all while you keep shipping code.

Your CLI Can Make Product Videos Now

You just shipped a feature. Now you need a demo video for the changelog, a social clip for Twitter, and a walkthrough for the docs. That is at least an hour of editing work you would rather spend coding.

What if you could tell an AI agent in your terminal to handle all of it?

With CLI-based AI agents like Claude Code, you can delegate video creation tasks to an agent that understands your project, drives Highlight Studio through its tools, and delivers polished videos while you keep working on the next feature.

How It Works

Modern AI coding agents run directly in your terminal and can interact with external tools through MCP (Model Context Protocol) servers. Highlight Studio exposes its project and editing capabilities through integrations that these agents can use.

Here is the workflow:

  1. You record your screen - Capture a raw screen recording of the feature you just built. Do not worry about pacing, mistakes, or polish. Just walk through the feature naturally.
  2. Hand it to the agent - From your CLI, tell the agent what you need: "Create a 30-second product demo from this recording. Add zoom effects on the key interactions, brand it with our intro and watermark, add subtitles, and export at 1080p."
  3. The agent edits for you - The agent opens your Highlight Studio project and works through the editing steps: trimming dead air, adding zoom regions at click points, applying your brand kit, generating subtitles via AI transcription, and kicking off the export.
  4. You get the finished video - The agent notifies you when the export is done. You review the result and share it.

What Agents Can Do in Highlight Studio

An AI agent with access to Highlight Studio's tools can handle most of the editing workflow:

  • Timeline editing - Split clips, remove sections, apply ripple deletes to tighten the pacing.
  • Zoom effects - Accept automatic zoom suggestions or create manual zoom regions with custom timing and easing.
  • Speed adjustments - Speed up boring parts (loading screens, navigation) and slow down key moments.
  • Subtitles - Trigger AI transcription, review the output, and style the captions.
  • Branding - Apply your Brand Kit, add watermarks, set intro and outro templates.
  • Canvas styling - Set aspect ratio, background gradients, padding, border radius, and shadows.
  • Annotations - Add text callouts, arrows, blur regions over sensitive content, and highlight boxes.
  • Export - Choose format (MP4 or GIF), quality level, and start the render.

Real Examples

Quick Changelog Demo

You just merged a PR. From your terminal:

"Take my latest recording, trim it to the key feature interaction, add auto-zoom on clicks, apply our standard brand preset, and export as a 720p MP4 for the changelog."

The agent handles the edit in the background. Two minutes later, you have a polished 20-second clip ready to embed.

Social Media Batch

You need the same demo in multiple formats:

"From this recording, create three exports: a 16:9 MP4 for YouTube, a 9:16 version for Instagram Reels with large subtitles, and a GIF under 10 seconds for Twitter."

The agent creates each variant with the appropriate aspect ratio, subtitle sizing, and format. You get three files ready to post.

Documentation Walkthrough

Your new API endpoint needs a visual guide:

"Edit this recording into a step-by-step tutorial. Add zoom effects when I interact with the code editor and the API response. Add text annotations labeling each step. Use slow motion on the response section. Export at source quality."

The agent adds zoom regions, places text annotations at the right timestamps, applies speed effects, and exports a detailed walkthrough.

Why This Matters for Developer Teams

Product videos are one of the highest-impact, lowest-priority tasks on most teams. Everyone agrees they need them, but nobody has time to make them. The editing workflow is the bottleneck - not the recording.

By delegating the editing to an AI agent:

  • Engineers stay in flow - No context switching between code and video editing. Record, delegate, keep coding.
  • Videos ship with features - Instead of "we will make a demo later" (which never happens), the demo is ready when the feature ships.
  • Consistency is automatic - The agent applies the same Brand Kit, style presets, and quality settings every time. No more inconsistent one-off videos.
  • Iteration is fast - Need to re-edit? Just tell the agent what to change. No need to relearn the timeline or redo manual work.

Getting Started

  1. Record your screen in Highlight Studio as you normally would. Focus on the feature, do not worry about polish.
  2. Open your CLI agent (Claude Code or any MCP-compatible agent).
  3. Describe what you want in plain language. Be specific about format, duration, and style if you have preferences.
  4. Let the agent work while you move on to the next task.
  5. Review and share the finished export.

The era of "I do not have time to make a demo video" is over. Your CLI agent has time, and Highlight Studio gives it everything it needs to deliver professional results.

ai agents · cli · automation · product videos · developer tools · featured