All pipelines
production
Screen recording
v2.1

Screen Demo

Screen recording pipeline. Two production modes: 1. REAL CAPTURE — takes raw screen capture (app demo, browser walkthrough, coding tutorial) via screen_recorder, cap_recorder, or playwright-recording and produces a polished demo video with callouts, zoom crops, subtitles, and cleaned audio. 2. SYNTHETIC (Remotion TerminalScene) — for CLI/terminal/install-flow demos, renders a deterministic terminal animation with typed commands, scrolling output, floating command pills, and blinking cursor. See .agents/skills/synthetic-screen-recording/SKILL.md. Preferred when the demo content is predictable (install walkthroughs, make targets, git clone, API key config) because it's faster to iterate, privacy-safe, pixel-perfect, and frame-accurate to narration cues. The idea-director picks the mode at brief time based on whether the demo surface is a real app UI (real capture) or a scriptable terminal session (synthetic).

Stages
7
Human approval gates
5
Default budget
$1.00
Recommended playbooks:
Minimalist Diagram
Clean Professional
Production modes
  • Real Capture — OS screen recording of a real desktop/browser session
  • Synthetic Terminal — Remotion TerminalScene — synthesized terminal animation

Stages

The agent reads each stage's director skill before working, then writes a checkpoint. Stages marked pause for your approval before the production continues.

  1. 1

    Idea

    Human approval

    openmontage/skills/pipelines/screen-demo/idea-director.md

    Produces

    • brief
    • decision_log

    Tools

    None — agent reasoning only

    Review focus

    • Brief accurately describes what the screen recording demonstrates
    • Target platform and duration are realistic
    • Callout strategy identified (what to highlight)

    Success criteria

    • Schema-valid brief artifact
    • Brief references the provided screen recording footage
  2. 2

    Script

    Human approval

    openmontage/skills/pipelines/screen-demo/script-director.md

    Produces

    • script

    Tools

    • transcriber
    • scene_detect

    Review focus

    • Transcription captures any voiceover accurately
    • Key UI actions are annotated with timestamps
    • Sections align with logical workflow steps

    Success criteria

    • Schema-valid script artifact
    • Script sections have timestamps from transcript or manual annotation
  3. 3

    Scene Plan

    Human approval

    openmontage/skills/pipelines/screen-demo/scene-director.md

    Produces

    • scene_plan

    Tools

    • frame_sampler

    Review focus

    • Zoom-crop regions identified for key UI moments
    • Callout overlay positions don't obscure important UI
    • Pacing allows viewer to follow the workflow

    Success criteria

    • Schema-valid scene_plan artifact
    • Every key workflow step has a corresponding scene
  4. 4

    Assets

    Human approval

    openmontage/skills/pipelines/screen-demo/asset-director.md

    Produces

    • asset_manifest

    Tools

    • subtitle_gen
    • tts_selector
    • image_selector
    • diagram_gen
    • audio_enhance

    Review focus

    • Subtitle file exists and matches speech timing
    • Reusable callout overlays (arrows, highlights, masks) are prepared
    • Audio cleaned of keyboard noise and background hum

    Success criteria

    • Schema-valid asset_manifest artifact
    • All referenced asset files exist on disk
  5. 5

    Edit

    Auto checkpoint

    openmontage/skills/pipelines/screen-demo/edit-director.md

    Produces

    • edit_decisions

    Tools

    None — agent reasoning only

    Review focus

    • Zoom-crop notes are smooth and minimal (no constant motion)
    • Dead time (loading, typing pauses) trimmed or sped up
    • Callout timing matches narration

    Success criteria

    • Schema-valid edit_decisions artifact
    • All cuts reference valid source files and assets
  6. 6

    Compose

    Auto checkpoint

    openmontage/skills/pipelines/screen-demo/compose-director.md

    Produces

    • render_report
    • final_review

    Tools

    • video_compose
    • hyperframes_compose
    • audio_mixer
    • video_trimmer
    • audio_enhance

    Review focus

    • Output video is playable and crisp at target resolution
    • UI text remains readable after crops and scaling
    • Audio is clear with no keyboard/click noise
    • render_runtime in edit_decisions matches proposal_packet — silent swap is a CRITICAL governance violation
    • For synthetic UI/terminal demos, render_runtime='remotion' (TerminalScene) is preferred; render_runtime='hyperframes' is valid for custom HTML UIs — in that case, lint+validate must pass before render

    Success criteria

    • Schema-valid render_report artifact
    • Output file exists and passes ffprobe validation
  7. 7

    Publish

    Human approval

    openmontage/skills/pipelines/screen-demo/publish-director.md

    Produces

    • publish_log

    Tools

    None — agent reasoning only

    Review focus

    • Metadata includes relevant software/tool keywords
    • Chapter markers align with workflow steps
    • Export package is complete

    Success criteria

    • Schema-valid publish_log artifact
    • Export directory contains video, metadata, and thumbnail concept

Manifest: openmontage/pipeline_defs/screen-demo.yaml. Director skills live under openmontage/skills/.