Uniform blog/Scout just got its biggest upgrade yet

Scout just got its biggest upgrade yet

Lars Birkholm PetersenPosted on May 21, 2026

6 min read

Scout just got its biggest upgrade yet

TL;DR

Scout, Uniform's agentic AI, has received its largest upgrade yet, introducing powerful features like customizable skills, vector search, bulk operations, and integration with third-party MCP servers. These enhancements make it easy for users to manage, automate, and personalize digital experiences at scale for both humans and AI agents. The upgrade empowers teams to work faster and smarter—driving efficiency, consistency, and innovation throughout their digital experience workflows.

Scout, Uniform's agentic AI, first launched in 2024. Over the past couple of years, it has evolved from a focused assistant into a deeply integrated partner across the entire Uniform platform, creating content, optimizing experiences, configuring personalization, and working alongside your team in real time.

At a rapid pace, AI is changing the way customers interact with brands. Uniform makes it possible for humans and agents to work together on creating the best experiences for both humans and agents.

Today, we're releasing the most significant set of enhancements since Scout's debut to make it easy for both humans and any agent to manage experiences. Here's what we're most excited about.

Skills: teach Scout how your team works

Skills let you extend Scout with your own expertise. Think of SKills as reusable instructions that encode your team's best practices, brand standards, and workflow preferences, so Scout doesn't just generate content; it generates your content and experiences.

A few examples of what skills can do:

Write in a specific voice. Describe your brand's tone, provide examples of writing you love, and Scout will match that style whenever you prompt it. Ask Scout to "write like our editorial voice" or "draft this in the style of our product team," and the output reflects your standards, not generic AI copy.

Apply your design system. Create a skill that maps out which components from your library are best suited for different page types, landing pages, category pages, event pages, SEO pages, and so on. When you ask Scout to build a full page, it selects the right components and writes the content to fit, all from a single prompt.

Run GEO, AEO, and SEO audits. Define what a thorough audit looks like for your organization, and Scout will evaluate content against those criteria. Better yet, once Scout identifies what needs improving, you can ask it to make the changes and audit in one workflow.

Enforce brand compliance. Codify your brand guidelines into a skill and let Scout check content against them before anything goes live.

Skills bring Scout closer to the way your team actually works. Instead of adapting your prompts to the AI, you teach the AI to adapt to you.

Connect with third-party MCP servers

Scout can now connect to any third-party MCP server directly from Uniform, opening up workflows that span your entire tool ecosystem.

Take project management as an example: connect Linear, and your team can see tickets across different states without leaving Uniform. A content editor can pick up a ticket, ask Scout to create the entry or experience it describes, and move the ticket to "Done" in Linear once it's been approved, all from within the Uniform interface.

However, this isn't limited to Linear. Any tool that exposes capabilities through an MCP server becomes accessible from Uniform. Connect planning tools to surface relevant conversations while creating content. Bring in data from analytics platforms to inform optimization decisions.

The key principle: any tool that is supported by a third-party MCP server can be called from within Uniform. Scout becomes the connective layer between your experience operations and the rest of your stack.

Vector search: find anything with natural language

Every entry, composition, and component in Uniform is now indexed for vector search. This means you can describe what you're looking for in plain language, and Scout will find it, even when exact keywords don't match.

Ask Scout to find all entries where countries in Northern Europe are mentioned, and it will surface content referencing Denmark, Sweden, Norway, Finland, and Iceland, without you needing to search for each term individually.

This is powerful on its own, but it becomes transformative when paired with bulk operations.

Bulk operations: delegate at scale

One of our favorite capabilities is how Scout can perform bulk operations autonomously. Instead of manually updating content one entry at a time, you describe the pattern and the action, and Scout handles the rest.

A few examples:

Update outdated references. "Find all blog post entries where 'best of 2025' is mentioned and update them to 'best of 2026'." Scout locates every match and applies the change across the board.

Fill in missing metadata. "Find all pages that are missing Open Graph data and generate it." Scout identifies the gaps, writes the metadata based on each page's content, and applies the updates.

Surface underperforming content. "Find all pages with low traffic and suggest improvements." Combined with Uniform Insights, Scout can analyze performance data and recommend specific optimizations, from SEO adjustments to content restructuring.

With bulk operations, you can orchestrate changes across your entire digital experience from a single conversation. You stay in control of how much autonomy Scout gets.

Autonomous mode vs. review mode

Every bulk operation and every individual action Scout takes can run in one of two modes.

Autonomous mode lets Scout execute changes directly, without requiring approval for each one. All changes are created in draft, so you always have the opportunity to review before publishing. This is ideal for high-confidence, repetitive tasks where you trust the pattern and want to move fast.

Review mode has Scout propose what needs to change and where, then present the full list for your approval. You review each suggestion, select which ones to proceed with, and Scout applies only those. This works well for sensitive content, brand-critical pages, or any situation where a human eye adds value.

You choose the mode that fits the task. Routine metadata updates can run autonomously. A content refresh across your highest-traffic pages may warrant review. The flexibility is yours.

Context-aware personalization

We've also deepened how Scout understands Uniform's personalization and experimentation capabilities.

You can now ask Scout to develop a personalization strategy based on your existing content and experiences. Scout will analyze what you have, then recommend which signals, enrichments, quirks, intents, and audiences should be created to support the strategy. Once you approve the recommendations, Scout will create them.

From there, you can ask Scout to tag content with the appropriate enrichments in a bulk operation, connecting your personalization taxonomy to your content library at scale, rather than entry by entry.

This turns what used to be a multi-sprint, cross-functional effort into something a single team member can accomplish in an afternoon.

We're excited about what these capabilities unlock for our customers and partners. Scout is no longer just an assistant that helps with individual tasks; it's becoming an autonomous teammate that can operate across your entire digital experience workflow.

Ready to try it? Get started with Scout or reach out to your Uniform account team to learn more.

Want to explore different use cases? Read this blog that Scout authored to learn more about the capabilities.