MCP Server: Connect AI agents directly to your content architecture
The conversation around agentic AI has moved quickly. The harder question for most enterprise teams isn’t whether to use AI agents; it’s how to give them structured, safe access to the systems that already hold your content.
Brightspot’s new MCP server answers that question. Built on the Model Context Protocol, it gives AI agents a standardized way to interact with your content management system: discovering content types, searching by keyword or semantic query, and creating or updating content — all within clearly defined boundaries you control.
That last part matters. Rather than opening the door and hoping for the best, Brightspot’s MCP implementation lets administrators scope agent access as they would a CMS user, limiting their access to specific sites, content types, or actions in the CMS. You decide what agents can and can’t touch. The result is a practical path to AI-assisted content operations that doesn’t require re-architecting your CMS or building one-off integrations for every tool in your stack.
The MCP server also positions Brightspot as a coordination hub for broader workflows — connecting to analytics platforms, marketing automation systems and digital asset management tools so agents can retrieve assets, update metadata and trigger actions without custom code. It works across Claude, ChatGPT, Copilot and any other MCP-compatible agent, so your investment in the integration compounds as your AI toolset grows.
Brightspot Toolkit Chrome extension: A shared window into every page you publish
Content teams spend a surprising amount of time on tasks other than editing: hunting for the right content record in the CMS back end, switching between tabs to check SEO metadata, manually running accessibility audits. The Brightspot Toolkit for Chrome is designed to eliminate that overhead.
The extension gives content, marketing and development teams a single panel that lives alongside any live page. With Click-to-edit, editors can jump from the page directly to its edit form in Brightspot — no searching, no manual configuration, no developer involvement. It’s the kind of feature that becomes a part of your every day and makes you wonder how you worked without it.
Beyond editing, the Toolkit brings together capabilities that most teams currently manage across separate tools: real-time SEO and GEO analysis that shows exactly what search engines and LLMs see at the moment content goes live; automated WCAG 2.1 AA accessibility audits; structured content inspection for developers; and Core Web Vitals diagnostics powered by Google PageSpeed Insights. New capabilities are added regularly and roll out automatically — no extension updates required.
For teams managing content operations across multiple roles and disciplines, having a shared, page-level view reduces the back-and-forth that slows publishing cycles down.
New plugin: Guided tours, built into the CMS
Onboarding is one of those problems that never fully goes away. New editors join. Workflows change. Organizations scale to dozens or hundreds of contributors across multiple sites. The standard tools (documentation, training sessions, third-party overlays) tend to create new dependencies rather than solve the underlying problem.
Brightspot’s CMS Tours plugin takes a different approach: guided, step-by-step onboarding built natively into the CMS itself. Admins can build tours without writing code, assembling steps in a drag-and-drop interface and scoping them by content type and user role. A new editorial hire sees the guidance relevant to their workflow. A marketing contributor gets a customized guide to that new page template when they need it. Everyone gets the right context at the right moment.
Tours launch automatically when a user opens a content type or navigates to an admin area for the first time. Returning users can restart them on demand. Full revision history means tours can be updated and improved with version comparisons and the ability to revert.
Because the plugin is native to Brightspot, it works seamlessly with your content models in a way that generic third-party tools simply can’t. For publishers managing large contributor bases, that context is what makes self-serve training actually work.
