I've been building a collaborative docs tool called Docules. The short version: it's a team documentation tool that doesn't have any embedded AI features. I use Claude code daily, but putting LLM’s into every workflow and charging for it is kinda insane. Every docs tool is adding AI auto-complete, AI summaries, "generate a page" buttons. Docules has an open API and ships an MCP server, so it connects to whatever you want to use LLM-wise. They can read, search, create, and edit documents through the API. The core product is just a docs tool that tries to be good at being a docs tool:
- Real-time collab with live cursors
- Fast — no embedded databases or heavy view abstractions slowing things down
- Hierarchical docs, drag-and-drop, semantic search
- Comments, version history, public sharing
- SSO, RBAC, audit logs, webhooks
Stack is React, Hono, PostgreSQL, WebSockets. The MCP server is a separate package so it's not coupled to the main app. I keep seeing docs tools bolt on half-baked AI features and call it innovation. I'd rather build a solid foundation and let you plug in whatever AI workflow actually makes sense for your team. Happy to answer questions about the architecture or the MCP integration.