How to Create AI Product Marketing Short Videos with short-video-factory (2026)

Short-form video has become the dominant format for product marketing. Platforms reward high-frequency posting, and production teams that can't keep up lose visibility to competitors who can. The bottleneck is no longer creativity — it's throughput.

That's why short-video-factory has been gaining traction fast. As of May 2026 it sits at nearly 4,000 stars on GitHub (YILS-LIN/short-video-factory), making it one of the most-watched open-source AI marketing tools on the platform right now. The pitch is simple: one-click generation of product marketing and general-content short videos, with AI batch automatic clipping, wrapped in a cross-platform desktop application.

This guide walks through how to make that promise work for a real marketing workflow.

What short-video-factory Actually Does

Before diving into steps, it helps to understand the two distinct jobs the tool handles:

1. One-click video generation for product marketing

The tool is built around the idea that you feed it product content and it handles the assembly. Rather than manually cutting clips, adding transitions, syncing audio, and sizing for different platforms, the AI pipeline does this in a single operation. The "one-click" framing refers to initiating this automated assembly — the human work shifts from editing to curation and prompt design.

2. AI batch automatic clipping

For teams producing volume — multiple SKUs, weekly campaign refreshes, social calendars — batch clipping is where the time savings compound. Instead of processing videos one at a time, the engine can handle multiple inputs in a single run. This is the feature most relevant to marketing ops teams and agencies managing several clients or product lines simultaneously.

The project is distributed as a cross-platform desktop tool, so Windows, macOS, and Linux users all have access without needing a cloud subscription or browser-based account.

Prerequisites

Before you start, you'll need:

• The short-video-factory desktop application, downloaded from the GitHub releases page
• Your product assets: footage clips, product images, voiceover audio, or scripts — whatever source material your campaign uses
• A clear sense of the output target: which platform(s) you're cutting for, and what aspect ratios and durations they require

Since this is open-source software distributed through GitHub, installation follows the standard pattern for the tool's platform. Check the repository's README for the latest build instructions, as these can change between releases.

Step 1: Organize Your Source Material

The AI can only assemble what you give it. Before opening the tool, structure your assets in a way that maps to your campaign:

• Product clips: raw footage showing the product in use, close-ups, lifestyle context
• Copy/script: the key messages you want the video to communicate — features, benefits, calls to action
• Brand assets: logos, color palettes, font files if the tool supports them

Good input organization pays off especially in batch mode. If you're generating videos for ten products, having a consistent folder structure per product makes it straightforward to queue them all in a single batch run rather than manually loading each one.

Step 2: Configure the Generation Parameters

Once your assets are loaded, you'll define the parameters that guide the AI assembly. Based on the tool's description, the core decisions at this stage are:

• Content type: product marketing versus general content. Product marketing mode is optimized for commercial output — the AI understands the goal is to show a product favorably and drive action. General content mode is better suited for brand storytelling, educational content, or top-of-funnel awareness pieces.
• Output format: short-form video formats vary by platform. A TikTok-optimized cut has different requirements than a YouTube Shorts clip or an Instagram Reel.
• Batch settings: if you're running multiple products or variations, configure the batch inputs here before initiating the run.

The "one-click" aspect means that once these parameters are set, the AI handles the editing decisions — clip selection, sequencing, pacing, transitions — automatically.

Step 3: Run One-Click Generation

With parameters set, trigger the generation. The AI batch automatic clipping engine processes your inputs and assembles the output video(s).

What the AI is doing under the hood, based on what the tool description tells us:

• Clipping: selecting the most relevant segments from your raw footage
• Sequencing: arranging clips in a logical narrative order for the product marketing context
• Assembly: combining clips into a complete video output

For a batch run with multiple products, all of this happens in parallel rather than sequentially, which is where the throughput advantage over manual editing becomes significant.

Step 4: Review and Iterate

AI-generated content is a starting point, not a finished asset. After the generation completes:

1. Review each output against your campaign brief. Does the pacing feel right? Does the clip selection show the product well?
2. Flag outputs that need adjustment — whether that's a different clip selection, different sequencing, or a full regeneration with modified parameters.
3. Iterate on the ones that are close. Since generation is fast, it's often more efficient to tweak parameters and regenerate than to manually edit the AI's output.

The goal of this review loop is to converge on a set of outputs you can confidently publish, not to achieve perfection on the first pass.

Step 5: Export and Distribute

Once you're satisfied with the outputs, export for distribution. Match your export settings to the platform requirements:

Export in batches where possible to keep the distribution workflow as efficient as the production workflow.

Extending Your AI Marketing Stack

short-video-factory handles the video production layer. For a more complete AI marketing workflow, two other trending GitHub projects cover adjacent tasks:

Affitor/affiliate-skills (403 ★, updated May 18 2026)

Affitor/affiliate-skills provides 50 AI agent skills specifically for affiliate marketing. According to the repository description, the skill set covers the full flywheel: researching trending content, writing data-backed posts, generating infographics, building landing pages, and deploying — with social intelligence built in. It works with Claude Code, ChatGPT, Gemini, Cursor, Windsurf, and other AI environments.

If your product marketing extends into affiliate channels, this skill pack can handle the written content and landing page layer while short-video-factory handles the video layer.

BrianRWagner/ai-marketing-claude-code-skills (296 ★, updated May 17 2026)

BrianRWagner/ai-marketing-claude-code-skills describes itself as "marketing frameworks that AI actually executes," designed for Claude Code. Where affiliate-skills focuses on affiliate-specific tasks, this project is focused on providing reusable marketing frameworks as executable AI instructions — strategy and campaign structure, not just content generation.

Together, these three open-source projects cover a meaningful portion of the modern marketing stack:

Who This Workflow Is For

This setup makes the most sense for:

• Solo marketers and freelancers who need to produce high video volume without a production team
• Small agencies managing multiple product clients who need batch throughput
• E-commerce operators running frequent SKU launches or promotional campaigns
• Developers building marketing tools who want to extend or contribute to open-source AI marketing infrastructure

The open-source nature of all three tools means there's no recurring SaaS cost — the trade-off is that installation and maintenance require more technical comfort than a plug-and-play cloud tool.

Summary

short-video-factory is the highest-starred AI marketing tool in the current GitHub trending data, and its core proposition — one-click product marketing video generation with AI batch automatic clipping — addresses a real production bottleneck for marketing teams. The cross-platform desktop distribution means it's accessible without cloud dependencies.

The workflow is: organize assets → set generation parameters → one-click batch generate → review and iterate → export and distribute. Pair it with Affitor/affiliate-skills for written content and landing pages, and ai-marketing-claude-code-skills for campaign framework execution, and you have an open-source AI marketing stack covering video, affiliate content, and campaign structure.

All three repositories are actively maintained as of May 2026 — check their respective GitHub pages for the latest releases and documentation before starting.