MeshWork is a creative workspace built around notes, discussions, AI, and Git sharing.
This manual will help you systematically understand how to use MeshWork.
Knowledge Cube
Start with MeshWork's knowledge model
MeshWork does more than store documents or draw a knowledge graph. It brings the discussion process, relationship network, and history of change together so every important topic can answer how it formed, what it connects to, and why it changed.
Questions, viewpoints, evidence, objections, and revised judgment are part of knowledge because their order and causality matter.
02
Graph: knowledge enters a network
Document references, topic branches, backlinks, and public libraries keep conclusions connected and discoverable for future work.
03
Cube: knowledge keeps its trajectory
A Topic Knowledge Card keeps the current state, while Git file history shows how it changed, turning a snapshot into a traceable trajectory.
Message flow = process · Knowledge Card = current summary · Git history = how the summary changed
Core Concepts
∞
Workspace
The main usage environment users see when entering MeshWork. It's a product and interaction layer concept, not a fixed disk directory or Git repository. The workspace is responsible for combining one private library and multiple shared spaces.
📁
Private Library
The data boundary belonging to the current user. Suitable for long-term personal knowledge and content that should not automatically enter team sharing. Markdown, PDF, and Word files can all appear in the tree as knowledge documents.
📄
Knowledge Document
A first-class document inside a knowledge base. Markdown is editable; PDF and Word are read-only, searchable, referenceable by AI, and can be converted to Markdown when needed. Scanned PDFs can use local Apple Vision OCR when embedded text is insufficient.
👥
Shared Space
A team or project-level data boundary. Each shared space contains a public library, discussion area, shared attachments, and can optionally have its own Git history.
💬
Discussion Area
An asynchronous discussion space within a shared space, supporting topic management, @ mentions, # document references, / commands, and AI agent collaboration.
🤖
AI Agent
In MeshWork, AI is a participant rather than an automatic replacement. It can suggest, but will not modify your documents without explicit confirmation.
Product Architecture and Knowledge Flow
MeshWork adopts a "private-first, progressive sharing" design philosophy, where knowledge flows from personal creation to team collaboration, ultimately settling into reusable long-term knowledge.
Knowledge Flow Path: Create and accumulate content in your private library, with AI providing intelligent assistance. When team collaboration is needed, you can directly share content to shared spaces: discuss in the discussion area, then valuable content further settles into the public library as long-term knowledge. AI provides assistance at every step, but won't automatically share your private content.
Manual Overview
01
System Setup & Configuration
Understand the relationship between the workbench, private library, and shared spaces, then configure appearance, AI, document assistant, discussions, Git repositories, and network proxies as needed.
Relationship between workspace, private library, and shared spaces
Application settings, document assistant, and discussion settings
Remote Git repositories, SSH keys, and HTTPS tokens
The data boundary for team or project collaboration. Learn how to create, join, and share spaces, use public libraries and discussions, and collaborate through standard Git remotes.
Create, join, and share shared spaces
Collaborative editing in public libraries
Topic management and AI collaboration in discussion areas
Remote permissions, branch isolation, and sync conflict boundaries