kanwas-ai/kanwas

Kanwas — Shared context board for teams and agents

What it solves

Kanwas provides a shared, multiplayer workspace where human teams and AI agents can collaborate in real-time. It eliminates the context gap between users and AI by allowing both to work on the same documents, evidence, and decisions on a single canvas, making AI interactions transparent and traceable.

How it works

The platform uses a canvas-based interface where AI agent tool calls stream directly into a shared timeline. It is built on a Git-backed markdown filesystem, ensuring that all work is stored as markdown files with full version history and no vendor lock-in. A dedicated CLI tool allows users to sync their workspace with a local filesystem, enabling the use of local editors or integration with other coding agents.

Who it’s for

  • Founders: To manage fundraising, customer interviews, and hiring plans in one place.
  • Product Managers: To turn interview snippets and competitor research into traceable PRDs.
  • Developers: To convert specs and designs into implementation plans that can be pulled into a local repository.
  • Marketers: To plan launches and iterate on copy variants side-by-side.
  • Sales: To create reusable account boards for research and proposal drafts.

Highlights

  • Multiplayer AI Collaboration: Real-time shared context between teams and agents.
  • Git-backed Markdown: All data is stored as markdown for portability and version control.
  • Local Sync CLI: A command-line tool for pulling and pushing workspace content to local directories.
  • Transparent AI Workflows: Agent tool calls are visible to all team members in the shared timeline.

Related

  • Project
    kiwifs/kiwifs

    A markdown filesystem that makes documentation writable, searchable, and versioned for both AI agents and humans, featuring native MCP support and Git-based auditing.

  • Project
    Team-Commonly/commonly

    An open-source workspace for collaborating with AI agents from any runtime, providing them with persistent identity and shared project memory to eliminate redundant context-setting.

  • Project
    MrLesk/Backlog.md

    A Markdown-native task manager and Kanban visualizer designed to help humans review AI agent work through spec-driven development and small, verifiable tasks.

  • Project
    kgoedecke/doop

    An open-source multiplayer design canvas that allows humans and AI agents to collaborate live by streaming HTML designs into shared frames.

  • Project
    cabinetai/cabinet

    A self-hosted, local-first knowledge base and AI agent orchestration platform that stores all data as Markdown files on disk to provide persistent memory for AI teams.