New Face Viral Video Replication Workflow: Build Your Own AI Video Skill in One Click

Understanding Viral Video Logic in One Word

In cross-border e-commerce and short-video marketing, viral content tends to share a common underlying logic. One video distills this into a single word: motion.

Take beauty and skincare content as an example — a person continuously moving closer to the camera weaves together skincare, makeup, and outfit transitions into a seamless flow. The constant visual movement paired with rapid cuts creates a strong visual punch, naturally embedding the product into the entire "glow-up" journey. This deliberate camera work and pacing is precisely what makes many viral videos so hard to replicate.

Now, with AI tools, ordinary creators can build these effects themselves — effects that once required a professional production team. This article breaks down the Agent workflow on the New Face platform and shows how it turns "viral replication" into a reusable, automated process.

New Face Viral Replication Workflow: A Full Breakdown

The starting point is remarkably simple. Open New Face, click the "plus" icon next to the chat box, and upload two things: a reference video and a product image. Then select "Viral Replication" from the Skill menu, and you're in the core collaboration loop with the Agent.

From there, chat directly with the Agent. It will automatically handle two key tasks: deconstructing the original video and extracting key frames. This is the foundation of the entire replication process — by identifying the shot structure and rhythm of the original, the AI accurately understands exactly where the viral magic happens. If the extracted frames feel incomplete, you can ask the Agent to keep pulling more key frames until you're satisfied.

This conversational approach is far more intuitive than traditional video editing software. Creators don't need to understand timelines or cinematography terminology — just describe what you need in plain language, and the Agent understands and executes.

From Key Frames to Final Cut: The AI Video Generation Pipeline

Once the key frames are confirmed, the generation phase begins. The Agent references these frames to produce new images featuring the creator's own products. Worth noting: the on-screen talent can also be regenerated using the built-in image model, allowing creators to further adjust clothing, poses, and scenes so the visuals align perfectly with their brand identity.

After the images are approved, select the appropriate video model to bring the talent to life — moving in sync with the original video's actions and rhythm. New Face offers two generation strategies:

Faithful Replication Mode

If you want to stay as close to the original as possible, choose "Faithful Replication" mode. This preserves the original video's camera movement, pacing, and visual impact to the greatest extent.

Creative Reimagination Mode

If you only want to keep the structural framework while redesigning the talent and scenes from scratch, go with "Creative Reimagination" mode. This gives creators far more creative freedom and sidesteps the homogenization problem that comes with directly copying the original.

Node-Based Editing: Fix One Shot Without Rerunning the Whole Pipeline

One of the most practical design choices in this workflow is the canvas-based node editing system. When a particular shot doesn't look right, creators can edit that individual node on the canvas — no need to regenerate the entire video.

This has a massive impact on creative efficiency. One of the biggest pain points with traditional AI video generation is the all-or-nothing problem — tweaking one detail often means rerunning the entire pipeline, which wastes both time and compute. Node-based editing lets creators pinpoint and adjust specific moments, dramatically reducing the cost of iteration.

Additionally, the Agent is currently in a limited-time free phase. Video deconstruction, key frame extraction, and prompt refinement don't consume any credits — credits are only deducted when you actually generate images or video. This means the entire exploration and refinement stage is essentially free, making it very beginner-friendly.

Turning Your Workflow into a Reusable Skill

The most valuable part of this workflow is its reusability. Once a video is complete and the results are confirmed, creators can simply tell the Agent to package the finalized shots, prompts, and generation steps into a dedicated Skill.

For future videos of the same type, all you need to do is swap in a new product image to generate versions across different categories — clothing, beauty, fragrance sachets, and more. In other words, a single reference video can gradually become a reusable content template.

Saved Skills appear in the chat list, ready to be called up at any time. This "build once, reuse many times" philosophy is the core value of AI Agent workflows — it doesn't just complete a single task, it codifies the creator's experience and methodology into executable, automated capability.

Conclusion: AI Workflows Are Reshaping Content Production

New Face's approach reveals a clear trend: AI content production is shifting from "one-time generation" to "workflow accumulation." For cross-border e-commerce sellers and short-video creators, this means producing batch after batch of stylistically consistent, structurally polished viral content at minimal cost.

More importantly, the Skill mechanism transforms "personal experience" into "reusable assets," so creators stop reinventing the wheel with every project. As these Agent tools become more widespread, the barrier to content production will keep falling — and the competitive edge will shift from "can you make it?" to "can you make it more creatively?"