- Home
- Skills
- Image & Video
- Short-Video Pipeline: Topic → Script → Finished Cut ⭐
Works with the AI tools you already use
Short-Video Pipeline: Topic → Script → Finished Cut ⭐
An end-to-end production workflow to turn news topics into scripted, storyboarded, and QA-verified short-form videos.
$14.99
Short-Video Pipeline: Topic → Script → Finished Cut ⭐
Example session with this skill installed
Process today's news about the new open-source LLM release into a 45-second vertical video script and production plan. Focus on the benchmark scores.
- Read your context and instructions
- Compiled the short-video pipeline
- Generated the image
I have processed the news regarding the 'Llama-3-Ultra' release. The focus is on its performance against proprietary models.
Topic Brief: Llama-3-Ultra Release
- Core Angle: Open-source AI finally matching top-tier proprietary performance.
- Hook: A new open-source model just cleared the 90% mark on MMLU benchmarks.
- Status:
PENDING_HUMAN_APPROVAL
45-Second Production Script
| Segment | Voiceover | Visual Direction |
|---|---|---|
| Hook (0-5s) | "Open source just caught up. The new Llama-3-Ultra is officially here." | Fast cuts of code blocks and the model logo. |
| Context (5-15s) | "Meta just dropped their largest model yet, and the benchmarks are staggering." | Screen recording of the GitHub repo and star count. |
| Detail (15-30s) | "We're looking at a 91.2% on MMLU, beating GPT-4o in three out of five categories." | Split screen: Benchmark table vs GPT-4o scores. |
| Impact (30-40s) | "This means high-tier reasoning is now accessible for local deployment without the API costs." | Graphic showing "Cloud Costs" vs "Local Hardware." |
| Close (40-45s) | "Is this the end of the closed-source lead? Let me know your thoughts." | Close up of narrator or brand logo with 'Follow' CTA. |
Production Timeline & Captions
- 00:00-00:05: [TEXT: OPEN SOURCE CAUGHT UP] "Open source just caught up..."
- 00:05-00:15: [TEXT: LLAMA-3-ULTRA] "Meta just dropped their largest model yet..."
- 00:15-00:30: [TEXT: 91.2% MMLU / BEATS GPT-4o] "We're looking at a 91.2% on MMLU..."
Publishing Package
- Title: Open Source AI Just Hit a Massive Milestone
- Description: Llama-3-Ultra is here and the benchmarks are changing the game for local LLMs.
- Hashtags: #OpenSource #AI #Llama3 #TechNews #MachineLearning
- Format: 9:16 Vertical, 4K, 30fps.
Next steps
- Verify the specific MMLU sub-category scores mentioned in the script.
- Approve
Human Checkpoint 1to proceed to asset gathering. - Select B-roll footage for the local hardware segment.
short-video-pipeline-topic-script-finish.png
PNG · 1536×1024
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
What you get
About this skill
The problem
Short-form video production often fails because creators either spend hours on manual scripting and storyboarding or rely on fully automated tools that hallucinate facts and produce low-quality edits. Balancing editorial integrity with the speed required for daily content cycles is a constant friction point.
What it does
- Evaluates news beats and trends to select high-potential topics based on hook potential and visual feasibility.
- Generates spoken-word scripts structured for retention with Hook, Context, and Why It Matters segments.
- Builds granular production timelines with timecodes for voiceover, visuals, on-screen text, and transitions.
- Implements a mandatory two-stage human checkpoint system to ensure factual accuracy and brand safety before rendering or publishing.
- Produces a complete publishing package including titles, descriptions, hashtags, and metadata.
Why this beats prompting it yourself
Generic prompts often produce long-winded scripts that don't fit the 9:16 vertical format or ignore the technical requirements of a video editor. This skill enforces a specific production logic that treats video as a multi-layered timeline rather than just a block of text, ensuring every second of the script has a corresponding visual and caption plan.
Use cases
- Transforming daily tech or finance news into 45-second vertical social clips.
- Standardizing a production workflow for a team of video editors and scriptwriters.
- Generating production-ready storyboards from raw research documents or press releases.
- Ensuring factual compliance and attribution in high-stakes news content.
Known limitations
Requires human intervention at two specific checkpoints to move from draft to final render. It does not perform actual video encoding or file exports without external tool integration.
How to install
Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.
- 1
Download the ZIP
Free skills download straight away. Paid skills unlock right after purchase.
- 2
Unzip into your skills folder
Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.
- 3
Ask your agent to use it
Restart the agent if it was already running. It picks the skill up automatically - no config needed.
Skills folder by agent
Click the path to copy it. Create the folder if it does not exist yet.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
30-day refund
Not a fit? Get your money back
Trust & safety
Security scanned
Verified clean 6 days ago
- Passed all security checks, Safe to install