0.2 BetaAGPLv3 / ELv2
An open-source knowledge base you install and run yourself. Organize documents and sources into connected knowledge called Context, then connect your AI tools to read and edit it with you.
Incorporate beta user feedback so anyone can understand product value at a glance.
Structured from the actual dtxWiki interface
Development plan
Connect an LLM to continue Ask chats within dtxWiki
An AI agent that performs tasks using the knowledge base
Why dtxWiki?
Files and notes accumulate across apps, but AI cannot automatically discover your context and past decisions. You end up copy-pasting background materials and re-explaining everything. dtxWiki unifies your scattered materials and past work into a single Context, allowing AI to read it and carry it forward.
Connected Knowledge
As you capture your thoughts, referenced notes, cited sources, and related readings link together automatically. Understand the broader picture at a glance without building complex graph setups.
References & citations
Humans & AI
Every revision saved
Find needed context
Knowledge MapPreview
Knowledge Map is a Preview feature you can enable in Settings.
Shared with teamEdited today
Example connections to 2 related notes
Shared Knowledge
Connect your favorite AI tools to dtxWiki via open-standard MCP* (Model Context Protocol). AI can search and read Context written by humans, and author or edit notes within permitted boundaries.
Connects insights and rationale into Context.
Discovers, reads, and edits within permitted bounds.
*What is MCP?An open standard protocol that connects AI tools to discover, read, and write knowledge in dtxWiki.
Transparent AI Collaboration
Request work in your connected AI tool. Chats shows what the AI searched, read, and edited in dtxWiki.
AI Work Trace
3 related Context notes · Example result
“Having to re-explain the entire project background every time I switch AI tools was the most tedious part.”
＋ Set hero message to “Stop repeating yourself.”
＋ Linked Customer Interview Summary as related Context
Self-Hosted Knowledge Base
dtxWiki runs in environments you manage directly. Storage destinations and external AI connection scopes vary by setup, with clear defaults documented in release notes.
Install through PyPI or npm, then open dtxWiki in your browser.
When connecting AI tools, only explicitly permitted contexts are shared with external providers.
Supported features, known limitations, and release updates remain publicly accessible at all times.
FAQ
dtxWiki is an open-source knowledge base that connects documents, decisions, and past work. Humans and AI read and modify the same Context together.
Yes. Install v0.2.1 from PyPI or use the npm launcher, then open dtxWiki in your browser. v0.2 does not include desktop installers.
Install the PyPI package or use the npm launcher, then open dtxWiki in your browser. Check the installation guide for requirements and background worker setup.
Use the source and PyPI or npm package for the version listed in the release. Check the release notes for changes in each version.
Yes. Any AI tool supporting MCP (Model Context Protocol), such as Claude Desktop or Cursor, can connect to dtxWiki to find, read, and propose changes within permitted scopes.
No. Chats does not store raw dialogue logs. It only records tool execution steps performed on dtxWiki, such as searching, reading, and editing Context.
dtxWiki can run in your own self-managed environment. When you connect external AI providers, only content within your explicitly permitted scope is sent to that AI service.
Yes. Teams can share and collaborate on the same knowledge base. Specific collaboration features such as invitations and access controls are detailed in the release notes.
dtxWiki is available under your choice of AGPL-3.0-only or Elastic License 2.0. Personal, team, and commercial use must follow the terms of the chosen license. Read LICENSE for the terms and commercial licensing options. Hosting and external AI services may have separate costs.
Graph View and Knowledge Map are Preview tools for exploring how knowledge connects. dtxWiki centers on building Context, working with AI, and checking tool activity in Chats. Knowledge Map is a Preview feature you can enable in Settings.
Use Cases
Not just for software teams. Anyone building long-term, compounding projects can leverage the exact same workflow.
Connect papers, primary sources, research notes with math and code, and lecture slides into comprehensive reports and future publications.
Read the guideLink customer interviews, PRDs, experiments, and decision rationales so the next iteration starts with full context immediately.
Read the guidePreserve operational SOPs, support guidelines, and recurring decision standards so onboarding teammates and AI start on the same page.
Read the guideOrganize brand guidelines, product specs, FAQ answers, and operational know-how in one place to automate repetitive workflows.
Read the guideConnect character sheets, lore, plot arcs, and channel tone-of-voice to keep your storytelling coherent across long-running creative projects.
Read the guideLink syllabi, lecture materials, grading rubrics, and feedback so you never have to rebuild course context from scratch.
Read the guideImport Existing Files
Import PDFs, documents, images, and datasets into Library to reference them in Context. Convert them to clean, AI-ready Markdown whenever needed.
OPEN SOURCE · AGPLv3 / ELv2
Inspect the source code and run it in your own environment under the license you choose. Shape dtxWiki's future through ideas, documentation, and code contributions.
CONTRIBUTE
Share what works well and what needs improvement.
Open GitHub Issues for features you need.
Collaborate on code, translations, and documentation.
Get Started with dtxWiki
From installation to your first Context, find all the guidance you need in one place.