Create complete product launch films with nine shot types, editable scenes and landscape, portrait and square exports. Includes device reveals, focused product detail, feature sequences, saved project edits, captions and original pulse audio. Original Skillrove workflow.
Two actual product films · editable scenes · three formats
index.html + film-engine.jsproject.json + copied assetslandscape/ + portrait/ + square/: MP4, poster, contact sheetcaptions.srt + .vttstoryboard.csvsoundtrack.wav (when selected)render-evidence.json + audio-evidence.jsonhandoff.md + delivery-manifest.jsonProduct teams and creators turning reviewed screenshots into a launch or feature film.
No voiceover, live recording, generative characters or automatic publishing.
Capabilities: nine complementary shots — kinetic typography, geometric statements, whole-image screenshots, browser/phone reveals, focused image regions, feature grids, advancing workflows, documented comparisons and closing CTA. Three palettes or four supplied brand colors. Local editor changes copy, timing, transitions, sequence, brand, focus and audio settings; saved/downloaded projects are valid inputs for a fresh render.
Deliverables: editable HTML/CSS/Canvas source, normalized project and copied permitted assets; landscape 1920×1080, portrait 1080×1920 and square 1080×1080 MP4s; posters, per-scene images, contact sheets, SRT/VTT captions, storyboard CSV, local soundtrack when selected, render/audio evidence and delivery SHA-256 manifest. Preview and exporter use one deterministic engine.
Requirements and run: Python 3.10+, local Chromium, Pillow, fontTools and FFmpeg. Install requirements.txt and Chromium in a virtual environment; python scripts/build-film.py --input examples/launch.json --output ./sample-film --render. Select 24/30 fps, formats or smaller drafts. Optional --cut-duration 30|60|90 proportionally retimes the supplied story; unusable scene timings require reauthoring. No API key or paid generation service.
Validation: two real product stories rendered at full resolution in all three formats. Actual 60-second Skillrove launch and 30-second Client Proposal feature film with explicitly illustrative project fees. Each encoded scene was reconciled against its Canvas frame; full frame count, aspect ratio, fps, decoding and audio-track checks passed. Independent fictional feature announcement exercised saving, JSON download, changed-project rerender, deterministic repeat exports and malformed-input/output preservation checks.
Limits: 20–90 seconds, 3–12 scenes, 1.5–15 seconds per scene. Supplied screenshots and facts need review; small desktop text can remain small after scaling. Preview is silent. Exports support original procedural pulse music or permitted local 16-bit PCM WAV with credit. No narration, live recording, generative characters, 3D or automatic publishing. English/Latin fonts are checked; a licensed local TTF/OTF/WOFF can be supplied for other glyphs.
Version 1.3.0 expands a simple walkthrough into a complete editable product-film workflow. Earlier 1.0.0, 1.1.0 and 1.2.0 packages and legacy Pillow commands remain available. Existing purchases and active memberships receive the active package. Price remains US$4.99. Original Skillrove code is MIT licensed; bundled fonts retain OFL notices.
FROM PACKAGE TO FIRST RESULT
Build the fictional nine-shot practice brief, then author your own launch story. The public studio includes two actual product films: a 60-second Skillrove launch and a 30-second Client Proposal feature announcement.
These steps use a project folder. Open the project in your agent and run terminal commands from its root.
Package folder .agents/skills/product-demo-studio
Download and unzip the package into the folder above. Keep all its files together, with SKILL.md directly inside that folder.
Downloads require a verified account and purchase access. Choose a new folder if this Skill is already installed.
Read SKILL.md, review the generator and check its runtime requirements before running:
cd .agents/skills/product-demo-studio
python3 -m venv .venv
.venv/bin/python -m pip install -r requirements.txt
.venv/bin/python -m playwright install chromium
.venv/bin/python scripts/build-film.py --input examples/launch.json --output ./sample-film --renderServe sample-film through a local static server. Play all three actual MP4s; select shots, apply edits, reorder scenes, change branding/audio settings and save the project. Download edited-project.json beside this folder's assets and run the builder against it into a fresh delivery. Check captions, product crops and the final CTA. Editor edits leave existing MP4s unchanged.
Python 3.10+; local Chromium, Pillow and FFmpeg. Setup is shown for macOS/Linux; use .venv\Scripts\python.exe on Windows. Nine shot types, three palettes, 20–90 seconds, 3–12 scenes, 24/30 fps and optional proportional 30/60/90-second retiming. English fonts are bundled with OFL notices; --font accepts a permitted TTF/OTF/WOFF. Preview is silent; actual exports can use original pulse audio or a permitted local PCM WAV. No narration, live screen recording, 3D generation, paid API or publication.
If sample-film already exists, change the output to ./sample-film-2. Existing files are never overwritten.
Use Product Demo Studio at .agents/skills/product-demo-studio. Read SKILL.md and review scripts/build-film.py. Run examples/launch.json into a new output folder and show me all generated files. Keep fictional sample content clearly labeled.
Tried this workflow? Help others decide with an honest review.
The listed individual package price is USD 4.99. Premium downloads require a confirmed purchase or active membership. Your agent subscription, model access and external provider credits are separate.
This release lists Codex, Claude Code, Cursor as intended installation hosts. Check the package’s dependencies and permissions before running it. These labels do not certify that every workflow has been tested in every agent.
The documented 1.3.0 workflow lists: index.html + film-engine.js; project.json + copied assets; landscape/ + portrait/ + square/: MP4, poster, contact sheet; captions.srt + .vtt; storyboard.csv; soundtrack.wav (when selected); render-evidence.json + audio-evidence.json; handoff.md + delivery-manifest.json. Some files or exports are optional; follow the package’s setup and inspect the delivered result.
Python 3.10+; local Chromium, Pillow and FFmpeg. Setup is shown for macOS/Linux; use .venv\Scripts\python.exe on Windows. Nine shot types, three palettes, 20–90 seconds, 3–12 scenes, 24/30 fps and optional proportional 30/60/90-second retiming. English fonts are bundled with OFL notices; --font accepts a permitted TTF/OTF/WOFF. Preview is silent; actual exports can use original pulse audio or a permitted local PCM WAV. No narration, live screen recording, 3D generation, paid API or publication.
