LLM Tools|Index 02
OpenKnowledge: A WYSIWYG Markdown Editor with Integrated AI
A new open-source macOS app and CLI offers a collaborative, what-you-see-is-what-you-get Markdown editor with direct integrations for LLMs like Claude and Cursor.
- Via
- AITECH TOKYO Editors
- Dateline
- TOKYO, June 25, 2026
- Date
- June 25, 2026
- Time
- 6 min read
Source
Hacker News TopTagline
WYSIWYG Markdown editor with integrated LLM agents
Who & Why
For a Tokyo-based product manager or technical writer preparing detailed specifications, OpenKnowledge offers a collaborative Markdown environment with direct AI assistance for drafting and refining technical content.
vs. Existing
It competes with Obsidian for local Markdown editing, offering true WYSIWYG and deeper LLM integration, while challenging Notion AI for collaborative document creation with a stronger focus on code and technical specs.
Tokyo Take
OpenKnowledge presents a compelling solution for technical documentation in English-first teams. Its immediate utility for Japanese-language workflows or integration with common Japanese enterprise tools remains limited, suggesting a wait-and-see approach for broader adoption here.
OpenKnowledge is an open-source, WYSIWYG (What You See Is What You Get) Markdown editor designed for collaborative document creation, available as a macOS application and a command-line interface.
Developed by Nick and his team at Inkeep, the tool launched on June 25, 2026, aiming to provide a Google Docs-like experience for Markdown files, a common format in technical documentation and development workflows. It addresses a gap where existing tools like Obsidian lack a true WYSIWYG interface and seamless LLM integrations.
The core offering includes a macOS app with a file navigator, the visual editor, and a link explorer. For users preferring a terminal-first approach, an embedded terminal and CLI are also provided.
A key differentiator is its direct integration with desktop applications for leading LLMs such as Claude, Codex, and Cursor. These AI agents can open an OpenKnowledge editor within their embedded web browsers, enabling a side-by-side editing experience. This facilitates scenarios like AI-assisted spec writing and building an “AI Second Brain.”
Under the hood, OpenKnowledge leverages a modern web stack including Tiptap/ProseMirror for the editor, Electron for the macOS app, and yjs for CRDT-based collaboration. The architecture includes a bidirectional lossless conversion pipeline between ProseMirror’s ASTs and Markdown, along with a dual-observer CRDT to keep states synchronized.
Collaboration features, including undo/redo and version history, are powered by CRDT and Git, offering a “no-code” feel while maintaining data privacy. The entire project is open-source, with its codebase available on GitHub, encouraging community contributions.
Pricing for OpenKnowledge is straightforward: it is fully free, local, and open-source. This positions it as a compelling alternative to proprietary collaborative editors or more fragmented setups involving separate Markdown editors and AI plugins.
The tool competes with established local Markdown editors like Obsidian, but distinguishes itself with its true WYSIWYG interface and deep LLM integrations. For collaborative document creation, it offers a more code-friendly and AI-integrated alternative to general-purpose tools like Notion AI or Google Docs.
Adjacent Tools
LLM Tools
Patronus AI Creates Virtual Testbeds for Autonomous AI Agents
Patronus AI introduces a platform for stress-testing AI agents in simulated environments, aiming to uncover vulnerabilities before real-world deployment.
LLM Tools
Anthropic's Claude Gains Paid Subscribers, Challenges ChatGPT
Anthropic's Claude is reportedly attracting paid consumers, signaling a shift in the premium AI assistant market previously dominated by OpenAI's ChatGPT.
LLM Tools
BeSimple AI's Audio Data Focus Signals New Frontiers in Voice AI
A Y Combinator-backed startup, BeSimple AI, is building foundational audio AI capabilities, hinting at applications beyond conventional transcription and voice assistants.