← 开源
kanwas-ai

kanwas

Kanwas — Shared context board for teams and agents

ApplicationsProductivityResearch & knowledgeTypeScript
在 GitHub 打开
增长势头
+024 小时新增 Star0.0%
752
Star
102
Fork
+-1
本周
1
贡献者
创建于 2026-04-22 · 更新于 2026-10-13 · 今日第 14216 名
主要开发者
marek-vybiral

kanwas-ai/kanwas 排名第 14,216,有 752 个星标。

同类别比较

kanwas-ai/kanwas 排名第 14,216,此处有 8 个同类条目;MewCo-AI/mewco_ai_assistant_comm 排名第 14,211。

同类别比较
名称排名指标
clmnin/summarize.site14,198
语言:
JavaScript
星标数:
753
分叉数:
82
简介:
Summarize web pages using OpenAI ChatGPT
r1n7aro/Locus14,203
语言:
Rust
星标数:
753
分叉数:
79
简介:
The open source Unity Dev Agent
talkingwallace/ChatGPT-Paper-Reader14,205
语言:
Python
星标数:
753
分叉数:
106
简介:
This repo offers a simple interface that helps you to read&summerize research papers in pdf format. You can ask some questions after reading. This interface is developed based on openai API and using GPT-3.5-turbo model.
MewCo-AI/mewco_ai_assistant_comm14,211
语言:
Python
星标数:
752
分叉数:
117
简介:
枫云AI助手社区版
jart/emacs-copilot14,233
语言:
Emacs Lisp
星标数:
750
分叉数:
22
简介:
Large language model code completion for Emacs
YoungCan-Wang/WyckoffTradingAgent14,240
语言:
Python
星标数:
749
分叉数:
216
简介:
Open-source Wyckoff trading agent and AI stock screener for volume-price analysis, A-share screening, CLI workflows, and MCP tools.灵感来自秋生trader @Hoyooyoo
getclawe/clawe14,244
语言:
TypeScript
星标数:
749
分叉数:
83
简介:
Multi-agent coordination system: think Trello for OpenClaw agents.
manparvesh/yoda14,247
语言:
Python
星标数:
749
分叉数:
151
简介:
Wise and powerful personal assistant, available in your nearest terminal
关键参数与规格
关键参数与规格
参数值
语言TypeScript
类别Applications
子类别Productivity,Research & knowledge
关键词Context engineering, Document understanding, Knowledge base, Notetaking
星标数752
分叉数102
贡献者1
创建日期2026-04-22
更新日期2026-10-13
1 天星标数0
7 天星标数-1
排名14,216
上次排名14,080
排名变化-136
7 天星标排名17,637
上次 7 天星标排名17,591
7 天星标排名变化-46
总星标排名14,054
上次总星标排名14,024
总星标排名变化-30
README 摘录

Shared context board for teams and agents Try Kanwas for free at kanwas. Kanwas is a multiplayer workspace for AI work. Teams and an AI agent share the same documents, evidence, and decisions, with the agent's tool calls streaming into the same timeline everyone sees. ** Turn a fundraising deck, customer interviews, MVP spec, and hiring plan into one canvas where the agent helps across all of them. Less context to keep in your head, more output across many fronts. ** Drop

interview snippets, tickets, and competitor screenshots on a board; get a discovery readout and a PRD with every claim traceable to its source. ** Pull a PM's spec, designs, and research onto a canvas; work with the agent to turn it into an implementation plan with tasks and acceptance criteria. Then the markdown into your repo and hand it to Claude Code, Codex, or whatever coding agent you use. ** Plan a launch with positioning, messaging, asset list, and timeline. The agent

drafts copy variants you can compare side-by-side and iterate on with the team. ** A reusable account board: research, comms history, stakeholder map, and proposal drafts. Why teams use it **Shared context that compounds. ** Every decision and outcome makes the next board better than the last. ** Generate structured, execution-ready artifacts for every stage of the work. ** Git-backed markdown filesystem with full version history. Quickstart Prerequisites Docker + Docker Compose An Anthropic API key (and/or OpenAI API key

) Run it Open and you're in. For local development (hot reload, running services with ) and the architectural walkthrough, see . CLI is a command-line tool for syncing a workspace with your local filesystem. Useful when you want to edit notes in your editor of choice, bulk-import markdown you already have on disk, or script workspace access from CI or another agent. Install Authenticate Opens a browser tab to authorize the CLI. Edit a workspace locally After the first , the

directory is bound to that workspace via . Import markdown from disk Walks the source path, picks up every file (other files are skipped), preserves directory structure, and creates them in the target workspace. List / script All commands accept or to skip the interactive picker, which makes them safe to use from CI or wrapping agents. Source: · npm: Community Questions, ideas, want to chat with the team? 💬 Kanwas Kollective on Slack Contributing We'd love help. Open an issue before large changes

so we can align on direction. Run and the relevant package's lint before opening a PR. First-time contributors will be asked to sign our Contributor License Agreement. License Kanwas is licensed under the Apache License 2. Acknowledgements Kanwas stands on the shoulders of Yjs, BlockNote, AdonisJS, E2B, and many other great open-source projects.

常见问题

记录了多少星标和分叉?

kanwas-ai/kanwas 有 752 个星标和 102 个分叉。

仓库排名是多少?

kanwas-ai/kanwas 排名第 14,216。

记录的语言和类别是什么?

kanwas-ai/kanwas 有 752 个星标,语言为 TypeScript,类别为 Applications。

哪个仓库排名最接近?

kanwas-ai/kanwas 与 MewCo-AI/mewco_ai_assistant_comm 比较,后者排名第 14,211,有 752 个星标。

README

Shared context board for teams and agents

Try Kanwas for free at kanwas.ai

What's Kanwas?

Kanwas is a multiplayer workspace for AI work. Teams and an AI agent share the same documents, evidence, and decisions, with the agent's tool calls streaming into the same timeline everyone sees.

Who it's for

  • Founders. Turn a fundraising deck, customer interviews, MVP spec, and hiring plan into one canvas where the agent helps across all of them. Less context to keep in your head, more output across many fronts.
  • Product managers. Drop interview snippets, tickets, and competitor screenshots on a board; get a discovery readout and a PRD with every claim traceable to its source.
  • Developers. Pull a PM's spec, designs, and research onto a canvas; work with the agent to turn it into an implementation plan with tasks and acceptance criteria. Then kanwas pull the markdown into your repo and hand it to Claude Code, Codex, or whatever coding agent you use.
  • Marketers. Plan a launch with positioning, messaging, asset list, and timeline. The agent drafts copy variants you can compare side-by-side and iterate on with the team.
  • Sales. A reusable account board: research, comms history, stakeholder map, and proposal drafts. Each deal makes the template better.

Why teams use it

  • Shared context that compounds. Every decision and outcome makes the next board better than the last.
  • Canvas + agent on one surface. Work alongside AI over the same evidence, ideas, and trade-offs — transparent to everyone.
  • Sharp deliverables in minutes. Generate structured, execution-ready artifacts for every stage of the work.
  • Your files, your repo. Git-backed markdown filesystem with full version history. No vendor lock-in.

Quickstart

Prerequisites

  • Docker + Docker Compose
  • An Anthropic API key (and/or OpenAI API key)

Run it

git clone https://github.com/kanwas-ai/kanwas.git
cd kanwas

# Env files — fill in API keys, APP_KEY, etc.
cp .env.example .env
cp backend/.env.example backend/.env
cp yjs-server/.env.example yjs-server/.env
cp frontend/.env.example frontend/.env

docker-compose --profile app up

Open http://localhost:5173 and you're in.

For local development (hot reload, running services with pnpm dev) and the architectural walkthrough, see docs/SYSTEM_OVERVIEW.md.

CLI

kanwas is a command-line tool for syncing a workspace with your local filesystem. Useful when you want to edit notes in your editor of choice, bulk-import markdown you already have on disk, or script workspace access from CI or another agent.

Install

npm install -g @kanwas/cli

Authenticate

kanwas login

Opens a browser tab to authorize the CLI. Auth is stored globally in ~/.kanwas/config.json.

Edit a workspace locally

mkdir my-workspace && cd my-workspace
kanwas pull            # interactive picker, downloads files into the current directory
# ...edit files in your editor...
kanwas push            # uploads local changes back to the workspace

After the first pull, the directory is bound to that workspace via .kanwas.json. Subsequent pull / push reuse it automatically.

Import markdown from disk

kanwas import ./notes                       # interactive workspace picker
kanwas import ./notes --name "My Workspace" # non-interactive, by name
kanwas import ./intro.md --id         # single file, by ID
kanwas import ./notes --dest research       # place imports under a subfolder
kanwas import ./notes --overwrite           # replace files that already exist

Walks the source path, picks up every .md file (other files are skipped), preserves directory structure, and creates them in the target workspace.

List / script

kanwas workspaces             # list workspaces
kanwas workspaces --json      # JSON output for scripting
kanwas pull --id        # non-interactive: pin to a specific workspace
kanwas pull --name ""   # non-interactive: by exact name

All commands accept --id or --name to skip the interactive picker, which makes them safe to use from CI or wrapping agents.

Source: cli/ · npm: @kanwas/cli

Community

Questions, ideas, want to chat with the team?

Contributing

We'd love help. A few notes:

  • Read docs/SYSTEM_OVERVIEW.md for the mental model and project-specific gotchas (especially around Yjs/BlockNote — clone semantics and transactions matter).
  • Open an issue before large changes so we can align on direction.
  • Run pnpm format and the relevant package's lint before opening a PR.
  • First-time contributors will be asked to sign our Contributor License Agreement. The CLA bot will comment on your PR with a link and the signing phrase — you only need to sign once.

License

Kanwas is licensed under the Apache License 2.0.

Acknowledgements

Kanwas stands on the shoulders of Yjs, BlockNote, AdonisJS, E2B, and many other great open-source projects.

排名相邻
同类项目