kitchen-co-mcp
Enables AI agents to interact with Kitchen.co via the Model Context Protocol, supporting tasks, boards, conversations, files, and more through secure profile-based authentication.
README
Kitchen.co MCP (Unofficial)
Unofficial community extension by Dimy Osman.
Not affiliated with Kitchen.co. This is not an official Kitchen.co product. It is not endorsed, sponsored, or approved by Kitchen.co, 2create.io, or their affiliates. “Kitchen” / “Kitchen.co” identify the third-party API only — no partnership or rights are implied.
Connect Cursor / VS Code AI agents to your Kitchen.co workspace through the public Kitchen API and MCP.
You supply your own workspace URL and API token. This extension does not provide Kitchen accounts or Kitchen.co support.
Install
Recommended — Extensions marketplace
- Open Extensions in Cursor (or another Open VSX–compatible editor).
- Search for
Kitchen.co MCP (Unofficial)ordimy-osman.kitchen-co-mcp. - Click Install.
Listings:
After install
- Acknowledge the one-time unofficial notice (first run only).
- Command Palette → Kitchen.co MCP: Add Profile.
- Enter a profile name, your workspace URL (or slug), and API token from Kitchen → Settings → Developer → API Token.
- Optional: List Profiles to confirm the profile is ready (
durable mcp.json: yes).
Tokens stay on your machine (OS keychain + encrypted vault). They are never written into mcp.json as plaintext.
Alternative — Install from VSIX
If you prefer a local package: download a release VSIX from GitHub, or build one (see Development), then Extensions → ⋯ → Install from VSIX….
What you can do
Once a profile is connected, agents can use Kitchen MCP tools to:
- Read conversations, messages, files, folders, boards, invoices, clients, members, docs, milestones
- Create / update tasks and post conversation messages
- Call other same-origin Kitchen API paths via
kitchen_request
See SECURITY.md for how credentials are stored and what protections apply.
Commands
| Command | Purpose |
|---|---|
| About / Disclaimer | Unofficial status and legal notice |
| Add / Edit / Remove Profile | Manage workspace URL + API token pairs |
| List Profiles | Status (key stored, durable MCP, etc.) |
| Re-register MCP Servers | Re-sync if tools disappear |
| Test Connection | Quick API check |
| Show Output Log | Diagnostics (success is silent on startup) |
Settings
| Setting | Default | Purpose |
|---|---|---|
kitchenMcp.autoRegister |
true |
Sync MCP on startup |
kitchenMcp.persistToUserMcpJson |
true |
Keep durable entries in ~/.cursor/mcp.json |
kitchenMcp.writePlaintextEnvFile |
true |
Short-lived env files for Cursor spawn |
kitchenMcp.wipeEnvFilesOnDeactivate |
true |
Wipe those env files when the extension stops |
Startup success is silent; you only get a toast if something fails. Setup help appears once on first install.
Disclaimer & legal
- Unofficial / third-party — independent use of Kitchen’s public API.
- No rights from Kitchen.co — no trademark license, partnership, or official support.
- Support — use GitHub Issues only; do not contact Kitchen.co about this extension.
- Your credentials — you are responsible for tokens and for actions agents take with them.
- AS IS — MIT License; no warranties. Kitchen may change their API at any time.
See NOTICE.
Development
For contributors building from source:
npm install
npm run compile
npm run lint
npm run audit
npm run package # produces kitchen-co-mcp-*.vsix
Publishing notes for maintainers: docs/PUBLISH.md.
License
MIT © Dimy Osman — see LICENSE.
Kitchen.co and related marks remain the property of their respective owners.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。