Now we’re exploring the next step: making agents operate on documents without a browser involved.
The idea is a server-side system where agents can read, write, and transform documents in real time — essentially treating your document layer as a programmable data store that agents can interact with directly. Internally we’ve been joking about the name “real-time agentic document database”, because that’s the shape it’s starting to take.
I’m curious how others would think about this direction.
If you had a database where the fundamental unit wasn’t rows/objects, but documents with semantic structure that agents can mutate, what else could or should it do?
Where are the obvious footguns?
What boundaries would you want between agent autonomy and document integrity?
Would love to hear how you’d expand or constrain a system like this.
@tiptap/extension-drag-handle
@tiptap/extension-drag-handle-react
@tiptap/extension-drag-handle-vue-2
@tiptap/extension-drag-handle-vue-3
@tiptap/extension-emoji
@tiptap/extension-details-content
@tiptap/extension-details-summary
@tiptap/extension-details
@tiptap/extension-file-handler
@tiptap/extension-invisible-characters
@tiptap/extension-mathematics
@tiptap/extension-node-range
@tiptap/extension-table-of-contents
@tiptap/extension-unique-id
They are all MIT-licensed and available via NPM. Contributions are very welcome.Happy coding and enjoy your weekend!
We just shipped an upgrade to our DOCX import and export capabilities, including new extensions, new endpoints, better formatting handling, and custom node support. This may be for you if your product deals with Microsoft Word files in any way.
We rebuilt our DOCX handling pipeline to give developers more control and flexibility:
- Import endpoint: Converts .docx to Tiptap JSON, accurately handles complex formatting (lists, tables, inline styles).
- Export extension: Generates .docx from your editor content, including custom nodes.
- Custom node support: Map internal components (e.g. callouts, embeds) to standard DOCX structures.
- Style control: Define exactly how DOCX output should look to match your app’s UI.
- Frontend or backend: Run conversion wherever it makes sense for your setup.
- Image upload handling: We give you hooks to manage storage your way.
What's still on the roadmap:
1) Import and export of DOCX headers and footers
2) Import and export DOCX pagination
3) Import and export comments, version history, and suggestions
Developer Docs: https://tiptap.dev/docs/conversion/import-export/docx
Product Website: https://tiptap.dev/product/conversion
We’d love your feedback! If you’ve struggled with document conversion (or just want to nerd out a bit), let us know your thoughts and experiences.
We started Tiptap a few years ago because we were frustrated by the state of rich text editors for modern apps. Most were either too limited, too rigid, or tried to solve everything with a monolithic setup. We wanted something more composable and framework-friendly, so we built it.
We originally monetized by selling individual “Pro” extensions. That worked early on, but over time we realized these smaller pieces weren’t where our value was. They were common needs, but not really our moat. In the meantime we’re focusing on larger feature bundles (like collaborative editing, AI-assisted content, documents, etc) that solve complete problems and include backend infra, UI components, and integration logic.
At the same time, we’ve sunset the free tier of Tiptap Cloud (where those extensions were included) and switched to a time-limited free trial. That wasn’t an easy call, but it helps us keep things sustainable while still supporting the open source side of the project and build the ecosystem and platform around.
We chose MIT over more restrictive licenses because we don’t want developers to worry about using these extensions in commercial work or side projects. That freedom is important to us. We’ll keep our core editor and these extensions open. The paid features now live in well-defined bundles tied to real product outcomes.
Here’s what’s free and open:
- Details, DetailsContent, DetailsSummary
- Emoji
- DragHandle (React and Vue)
- FileHandler
- InvisibleCharacters
- Mathematics
- TableOfContents
- UniqueID
Happy to answer questions, or just hear what you think. Curious how other folks have handled licensing shifts like this, or if you’ve had similar tradeoffs in open source vs. business goals.Thanks, Philip
If you’ve ever tried wiring up AI inside a document editor, you’ve probably dealt with:
- Extracting context from complex document structures
- Handling prompt input + streamed output
- Supporting undo/redo for AI changes
- Designing UI for accepting/rejecting changes
- Multiplayer session state and conflicts
It’s a lot of work, and almost none of it is model-specific.This new Tiptap toolkit gives you a clean way to define AI Agents that can read and edit rich text based on user-defined tasks. You can trigger agents manually, automatically, or in response to structured input.
Works with OpenAI or your own backend + LLM stack. Built on top of the same multiplayer engine behind Tiptap’s collaboration features.
We also include an AI Changes extension so users can review and accept/reject generated edits, like a built-in code review for content.
There’s a live demo here: https://ai-agent.tiptap.dev/
Developer Docs: https://tiptap.dev/docs/content-ai/capabilities/agent/overvi...
Happy to answer questions or just hear what you’re building :-)