For Developers

Let the MCP drive.

FlowFrame turns screen recordings into narrated, edited videos. MCP lets Claude Code, Cursor, or Codex drive capture, narration, and export — so you ship the release tour, not another afternoon in the timeline.

Make your first video →

FlowFrame project gallery showing developer workflows: SDK quickstart, CLI demo, webhook integration walkthrough, v0.5.11 release notes tour

What if your release video was built by an Agent?

We built this walkthrough for ParseForge — a fictional open-source library. In Cursor, Claude Code, or Codex, one command to FlowFrame MCP: draft the tour, record the playground, export an MP4 for the GitHub release.

ParseForge isn't real; the workflow is. AI narration, MCP automation, typing animations, cursors and clicks, and polish from Director.

What you get

🤖

Hand the keys to your agent

One prompt → finished MP4. Cursor, Claude Code, or Codex.

💻

Capture what devs actually use

Code, terminals, simulators — all in one tour.

🎬

Brand every release the same

Reusable intro/outro snippets. Drop and ship.

📝

Ship video with every changelog

A 60-second tour from each release, auto-built.

Hand the keys to your agent.

The MCP server runs locally and exposes 107 tools across capture, steps, narration, snippets, export, and Director scoring. Same prompt, finished project.

Works with Claude Code, Cursor, and Codex. See the full tool list →

You: Open my latest screen recording, generate
     narration, add the team intro snippet, and
     export 1080p.

Claude: Importing capture…
        Drafting narration with your voice…
        Prepending intro-v3.snippet…
        Scoring with the Director (8.4/10)…
        Exporting MP4. Done.

Superpowers unlocked with your API keys

FlowFrame works without keys. Paste in the keys you already have and the app levels up — the Director, AI narration, licensed media, and agent automation all run on your providers. Keys stay on your machine.

OpenAI / Anthropic / Gemini

Director: scores every step red/yellow/green, suggests rewrites, can rewrite the entire tour from a brief.

Director + drafts

OpenAI

AI narration voices via OpenAI TTS. Generate the voiceover for a whole tour in one pass.

Narration

Pixabay & Freesound

CC0 imagery and Creative Commons music + SFX, searchable from inside the editor.

Media

YouTube

Import existing videos as source material — clip, narrate, restitch.

Source

Bring your own keys. Same app, more horsepower.

Your demo deserves a director.

Hand your release to Claude Code, Cursor or Codex and get a finished video back. Free for 7 days.

Make your first video →