Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
DeepSeek
Muse+22 moreClaude CodeCursorCodex CLIDeepSeekMuse
Dots
Manus
Grok
Grok Bot+18 agentsProduct Video Studio for Codex
Turn product pages, repositories, screenshots, and brand assets into reviewed promo videos with scripts, narration, subtitles, and local rendering.
$5
Product Video Studio for Codex
Example session with this skill installed
Use $enhe-product-video to create a 45–60 second landscape launch video for the product in the current repository.
Read the README, product screenshots, logo, and relevant source files as evidence. The audience is ordinary AI users, not developers.
First show me the complete narration, on-screen text, subtitles, call to action, and storyboard for approval.
Ask whether I want to provide a portrait for the cover and author ending.
Use natural standard-Mandarin male narration, transparent Chinese subtitles, and a clear website call to action.
Do not invent features, customers, metrics, reviews, or results.
Do not render until I approve the copy and representative frames.
After rendering, return the MP4, cover, subtitle file, script, storyboard, asset list, and quality report.
- Read your context and instructions
- Compiled the product video studio
ENHE Product Video Studio — Completed Demo
Production Summary
I reviewed the repository README, public product materials, approved showcase assets, and the product’s actual workflow.
The video was built for ordinary AI users who need to understand what the tool does, how the workflow works, and what they receive at the end.
| Item | Result |
|---|---|
| Product | ENHE Product Video Studio |
| Format | 1920×1080 landscape |
| Duration | 62 seconds |
| Language | Chinese |
| Narration | Synthetic Mandarin narration |
| Subtitles | On-screen subtitles |
| Main goal | Explain how scattered product materials become a reviewed promotional video |
| Call to action | Visit the ENHE website for more AI tools and solutions |
Approved Story
The video follows one clear message
A product may already be finished, but people still need a simple way to understand it.
Instead of starting with visual effects, the workflow first organizes the product story, confirms the script and storyboard, then produces and checks the final video.
Storyboard
Scene 1 — Opening
Question
“Your product is ready. What about its introduction video?”
The opening establishes the problem without making unsupported claims.
Scene 2 — The problem
Product notes, screenshots, brand files, and feature descriptions are often scattered across different places.
The viewer sees why a clear product story is needed before video production begins.
Scene 3 — The workflow
The video shows the main production path
Product materials
→ Product summary
→ Script and storyboard
→ User approval
→ Local animation and rendering
→ Quality review
Scene 4 — What the user receives
The video presents the actual deliverables:
- Promotional MP4
- Cover
- Narration
- Subtitles
- Script
- Storyboard
- Asset records
- Quality and render reports
Scene 5 — Review and reuse
The workflow records approved assets and production state so unchanged work can be reused when the project continues later.
Changed material is checked again instead of silently reusing an outdated result.
Scene 6 — Website and author ending
The final scene directs the viewer to the ENHE website and presents the approved author information.
Delivered Showcase Files
- Video:
docs/showcase/enhe-studio-demo-62s.mp4 - Cover:
docs/showcase/enhe-studio-cover.png - Supporting public showcase images:
docs/showcase/
The local production workspace also keeps the script, storyboard, video specification, asset records, and quality reports used for the task.
Review Status
The published showcase contains AI-assisted visuals and synthetic narration and was reviewed by the author before publication.
The example demonstrates a completed production workflow. A fresh public installation will not reproduce the exact same video automatically because private brand assets, local voice references, and local models are not included in the public repository.
Connects securely to your tools. The creator never sees your data.
About this skill
Make a product video people can understand
A finished product is hard to sell when its video only lists features or uses effects that do not explain anything.
ENHE Product Video Studio turns the product material you already have—such as a repository, product page, README, screenshots, logo, and approved brand assets—into a clear promotional video.
It starts with the product story
The workflow first works out:
- Who the product is for
- What problem it solves
- Which features are worth showing
- What evidence supports each claim
- What the viewer should do next
It does not invent customers, metrics, reviews, awards, results, or features that are not supported by the material you provide.
Review before rendering
You see and approve the complete narration, on-screen text, subtitles, call to action, and storyboard before the final render.
You can also review the cover, author ending, representative frames, visual direction, and narration separately.
This makes it easier to fix the message early instead of rebuilding an entire finished video.
What it can deliver
- An MP4 product introduction or promotional video
- A video cover and author ending
- Mandarin narration when the local voice setup is available
- On-screen subtitles and an SRT subtitle file
- A product summary
- Full narration and on-screen copy
- A scene-by-scene storyboard
- A video specification
- An asset list with source and usage information
- Render, quality, and delivery reports
Videos can be created in 1920×1080 landscape or 1080×1920 portrait format, with an approved duration between 8 and 90 seconds.
Built around real product evidence
The main video uses real product screens, documents, screenshots, workflows, and approved brand assets instead of generic feature lists or unsupported claims.
Each product statement should be connected to a real source, screenshot, file, or supplied asset.
Local production
Codex prepares the message, script, storyboard, and production files. HyperFrames handles local animation and rendering.
The workflow can reuse unchanged, approved work when a task continues later. If an asset, script, or scene changes, only the affected stages need to run again.
Why it is different
- The product story comes before the visual effects.
- Copy and storyboard are reviewed before rendering.
- Important claims must be supported by real evidence.
- Approved assets and results are versioned for later reuse.
- Subtitle, layout, sound, asset, and render checks are recorded.
- Private assets and voice references stay local by default.
- Paid cloud APIs and automatic publishing are not enabled by default.
Best for
- Independent developers
- One-person companies
- AI builders
- Product creators
- Open-source maintainers
- Small software teams
- Content creators introducing tools and workflows
- People who already have a product but need a clearer way to present it
Before you buy
The complete local workflow is currently verified on Windows x64.
You need Git, Node.js 24.15 or newer, npm, and a Codex environment that can read local files and use Skills.
Mandarin narration needs a separate local Python and voice-model setup. Private voice references, local models, private portraits, and the complete private brand library are not included.
This is not a browser video editor, cloud rendering service, bulk video factory, automatic social publisher, or voice-imitation tool.
Every final video still needs human review before it is published.
<img src="https://geeduzbsatvodnlijqfs.supabase.co/storage/v1/object/public/product-media/272dd91a-eb48-4b36-b5cd-2ad800a448da/content-images/1789401922655-yp0otaz3w1c.png" alt="ChatGPT Image 2026年9月11日 17_00_47 (2).png" loading="lazy" />Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Video Skills.
This is a local production workflow, not a browser video editor or cloud rendering service.
The complete workflow is currently verified on Windows x64 and requires Git, Node.js 24.15 or newer, npm, and a Codex environment that supports Skills.
Videos currently support 8–90 seconds in 1920×1080 landscape or 1080×1920 portrait format.
Mandarin narration requires a separate local Python and voice-model setup, plus human listening review. Voice models, private reference recordings, private portraits, and private brand assets are not included.
A fresh installation cannot reproduce the public 62-second showcase exactly because some approved private assets and local models are not distributed.
Current new videos focus on real product screens, documents, and workflow scenes. Built-in mascot characters are not a promised option for new jobs.
The workflow does not provide automatic social publishing, bulk unattended video production, paid cloud APIs, unrestricted web scraping, or imitation of a real person’s voice.
Final videos, claims, assets, narration, and publication rights still require human review.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 month ago
- Accounts
GitHub
- Software
Node.js
Permissions
Allowed Hosts
File Scopes
ENHE Product Video Studio reads only the product materials selected by the user, such as README files, source files, screenshots, logos, audio, and public product URLs. It runs local Node.js, HyperFrames, Chrome, FFmpeg, FFprobe, and optional Python narration tools. Project files, generated assets, caches, reports, and videos stay inside the local project workspace, mainly under `projects/`, `.tools/`, `.cache/`, and the current task directory. Network access is used for npm dependency installation, verified downloads of pinned local tools, optional local narration dependencies, and public URLs explicitly supplied by the user. The setup does not require administrator access and does not change the system PATH or Windows registry. Environment variables are used for local tool paths and to disable telemetry and update checks. The workflow does not request or read API keys by default. Private portraits, voice references, cookies, tokens, local models, and private brand assets are not uploaded or included in the public package. Paid APIs, cloud rendering, automatic publishing, and external uploads are disabled by default.