Content.One adds MCP server for AI-driven multi-site CMS workflows

Content.One said it has made its MCP server generally available, giving marketers and internal AI agents the ability to create websites, schemas, pages and CMS tasks with natural-language prompts. The launch targets federated organizations and promises faster production, tighter governance and less dependence on developers. Why it matters: - Content.One is betting that multi-site marketing teams will let AI handle more of the content and site-building work that usually slows launches. - The new MCP server is designed for franchises, multi-chapter nonprofits and multi-location brands managing hundreds of properties. - Content.One says the approach can compress enterprise workflows from 32 hours to one and turn ideas into live experiences faster. What happened: - Content.One announced the general availability of its MCP server on June 15, 2026. - The server lets non-technical marketers launch campaigns and entire websites with natural-language prompts. - Content.One said the functionality is available immediately to enterprise customers including Sony Electronics, The Salvation Army, Kin Insurance, Singlife and the Phoenix Suns. - The platform now exposes its APIs, content models and stateless infrastructure directly to AI agents through the MCP server, without an intermediate layer. - Marketers can use Claude or Gemini through the MCP server, or an organization can connect its own internal agents. The details: - The MCP server exposes accounts, content items, models, fields, audit logs, media bins, labels, settings and stylesheets to any MCP-compatible AI client. - Authentication is stateless and scoped by session token, so existing role and permission structures apply automatically. - Content.One says the platform supports roles, permissions and team-based governance across agentic workflows. - The new workflow can generate full content schemas in minutes from a natural-language brief. - The system can draft complete pages in roughly one to two minutes, including copy, co-author selection, brand-guideline application, image generation and page assembly. - Content.One can build HTML components from screenshots and generate matching components for templates. - Internal demo environments with full content, components and structure have been built from scratch in roughly 90 minutes. - Users can paste a Meta Pixel or other tracking script into chat and install it globally, on a single model or on a specific page. - The platform can upload, name, optimize and index images automatically when a user drops an image into chat. - The new SEO and GEO Analyzer agent scans pages for issues such as missing Open Graph tags and duplicate H1s, applies fixes in the live CMS and tracks scores over time. - Calendar, Workflows, Publisher, Forms, the Pop-up manager and the drag-and-drop builder all support natural-language control. - Multilingual prompting is supported in Spanish, Dutch and German. - The platform’s data architecture was originally built as Zesty.io and was designed for programmatic access from day one. - Security and governance features include role-based access controls, custom roles, SSO integration, full audit logging and permission cascades. - GDPR and CCPA requirements are handled at the platform level. - A globally distributed CDN supports 99.999% availability for high-traffic, multi-region properties. Between the lines: - Content.One is positioning the MCP server as a native extension of an already structured CMS, not as a bolt-on AI feature. - The company is also arguing that federated organizations need centralized governance and local autonomy at the same time. - The product framing suggests a push to move marketers from request-and-review workflows to prompt-and-approve workflows. - Randy Apuzzo said the agentic page creator cuts enterprise workflows down from 32 hours to one. - Apuzzo also said the company sees AI as “a world of action on your behalf” and described the platform as a way to combine insight, vendor tools and expert knowledge in one experience. What’s next: - Content.One is directing prospective customers to request a demonstration. - The company is likely to push adoption through existing enterprise accounts and multi-site organizations that want faster content production with tighter controls. - As more teams connect their own agents, the value of the MCP server will depend on how well organizations balance speed, governance and brand consistency. The bottom line: - Content.One is turning its CMS into an AI-operable control layer for complex, multi-site marketing operations.