The AI-native content engine. Framework-agnostic file-based JSON content, visual admin UI, AI agents, workflows, and a static build pipeline. Your content as flat JSON — render it with Next.js, Laravel, Django, C#, Rails, or anything that reads files.
Features
Content modeling, persistence, media pipelines, AI orchestration, and static output — all in one embeddable TypeScript library.
Provider-agnostic agents for content generation, rewriting, translation, and SEO optimization. Swap between Anthropic, OpenAI, or local models.
22 field types. Collections and blocks defined in TypeScript. Every piece of content is typed, validated, and introspectable.
9-phase build pipeline: HTML, sitemap, robots.txt, RSS, llms.txt, per-page markdown. Zero runtime JS unless you opt in.
First-class Next.js integration (App Router, Server Components, ISR). Also works with Astro, plain Node.js, or any static host.
Sharp-based image processing with AI-generated captions and alt text. WebP conversion, responsive variants, EXIF extraction.
Field-level protection. AI agents can never overwrite human edits. WriteContext actor threading through all CRUD operations.
Translation groups link documents across locales. AI auto-translates. Hreflang, locale routing, and language switcher built in.
Dual scoring: 13 SEO rules for search engines + 8 GEO rules for AI citation. Combined Visibility dashboard in admin.
Built-in AI chat that knows your content, schema, and configuration. Ask questions, get answers with citations from your own docs.
Every page in the CMS admin has contextual HelpCards— collapsible help panels that explain the feature you're looking at. Each HelpCard links directly to its expanded documentation page here.
The same content source powers both: in-app help is concise and actionable, while docs pages go deep with code examples, configuration reference, and guides.
Model Context Protocol
Two MCP servers. One public and read-only. One authenticated for content production.
Built with @webhouse/cms — dogfooding our own product
