v0.10.0 is now stable for release
- Thread & Academic Content Type
- Linked Knowledge between books
- Style References
- Worlds Search, you can now public your world settings
- Markdown & DOCX Import
v0.10.1 should be out in next two days.
Entity Cross-Navigation
Entity names mentioned in knowledge base card content are rendered as clickable links, click to jump to and expand the referenced entity
Guardian Jump-to-Issue
Click a guardian issue to jump directly to the problematic text in the editor, with character-offset-based position mapping for quick-check rules
v0.9.0 Stable -> After using Clawbot with MCP on Creader. All major P0 bugs are fixed
- Now support Gemini -> next up Kimi/MiniMax
- inline AI, ghost text and slash command
- Privacy statement on AI training
- Global search
- Upgraded the new homepage
Finalised the Creader v2 roadmap, from writing editor to world expression platform.
The current system works. But when I look at it closely, I see gaps everywhere, not bugs, but features waiting to happen.
Two ideas have occupied my mind recently: anticipation and Theory of Mind. Anticipation concerns the meanings and values a reader expects. Theory of Mind concerns predicting how readers interpret what is given.
AI is already great at analysis and logic. But writing isn't just logic. It's what the characters believe, what the reader expects, and what the author intends, all at once.
That's what v2 is building toward.
The roadmap:
- Foundation — codebase cleanup (done)
- MCP — your stories accessible from Claude, OpenClaw, any AI client
- Writing Core — inline AI v2, global search, referral
- Ecosystem — thread editor (Twitter publishing), academic mode, import
- AI Eval — measure every AI interaction, improve every prompt
- Story Structure — scenes, acts, corkboard view
- Integration — cross-platform publishing, narrative intelligence
The long bet: Narrative Intelligence, AI that understands character beliefs, reader expectations, and author intent simultaneously.
By v3, every book project gets its own dedicated model.
creader.io