Colab MCP
🪐 Colab MCP (Model Context Protocol)
Сервер MCP (Model Context Protocol), который бесшовно связывает вашего локального ИИ-агента с сеансом Google Colab, запущенным в браузере.
✨ Возможности
Подключает локальных ИИ-ассистентов напрямую к блокнотам Colab в браузере
Поддерживает выполнение кода Python в Colab через агента
Читает состояние блокнота Colab и взаимодействует с ним
Related MCP server: colab-mcp
💻 Поддерживаемые клиенты
Для работы этого MCP-сервера требуется клиент, поддерживающий notifications/tools/list_changed, который должен быть запущен локально на вашем устройстве.
Популярные клиенты, соответствующие этим критериям:
🚀 Установка и настройка
Установите
uv(чрезвычайно быстрый установщик и резолвер пакетов Python):pip install uvНастройте ваш MCP-клиент (например, в файле
mcp.jsonили аналогичном конфигурационном файле):{ "mcpServers": { "colab-mcp": { "command": "uvx", "args": ["git+https://github.com/googlecolab/colab-mcp"], "timeout": 30000 } } }Примечание для сотрудников Google (или тех, кто использует нестандартные индексы пакетов): Возможно, вам потребуется добавить
--index https://pypi.org/simpleв массивargs.
💬 Проблемы и обсуждения
Мы используем Discussions на GitHub в качестве основной площадки для обсуждения проблем и запросов на добавление функций.
По мере того как обсуждения перерастают в четкие задачи, сопровождающие будут преобразовывать их в отслеживаемые проблемы (issues). Этот рабочий процесс помогает нам гарантировать, что трекер проблем остается свободным от дубликатов, понятным и максимально ориентированным на действия.
⚠️ Пожалуйста, НЕ открывайте issues напрямую.
🤝 Участие в разработке
Хотя мы ценим интерес сообщества, в настоящее время у нас нет ресурсов для проверки внешних вкладов. Мы хотим избежать ситуации, когда Pull Requests от пользователей остаются без внимания, поэтому в данный момент мы не принимаем внешние вклады.
Если у вас есть отличная идея или вы столкнулись с проблемой, мы будем рады узнать об этом на нашей странице Discussions!
🛠️ Внутренняя информация (для разработчиков Colab)
Предварительные требования
Требуется
uv(pip install uv)Настройте git-хуки для запуска предварительных проверок репозитория:
git config core.hooksPath .githooks
Настройка локальной разработки (Gemini CLI)
Чтобы протестировать вашу локальную версию с помощью Gemini CLI, используйте следующую конфигурацию:
{
"mcpServers": {
"colab-mcp": {
"command": "uv",
"args": ["run", "colab-mcp"],
"cwd": "/path/to/github/colab-mcp",
"timeout": 30000
}
}
}MCP_Colab
Available Tools
1 toolopen_colab_browser_connectionA
Opens a connection to a Google Colab browser session and unlocks notebook editing tools. Returns a boolean representing whether the connection attempt succeeded
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Output Schema
| Name | Required | Description |
|---|---|---|
| result | Yes |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description must fully disclose behavior. It states the outcome (connection success) and return type (boolean), but lacks details on side effects, prerequisites (e.g., must be in Colab environment), or what 'unlocking' entails.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise, two sentences, front-loaded with the action and purpose. No wasted words.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given zero parameters and a simple return type, the description is mostly complete. However, it could mention prerequisites or environment requirements to fully prepare the agent.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema has no parameters, so description does not need to elaborate. The description adds value by explaining the boolean return meaning.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states it opens a connection to a Google Colab browser session and unlocks notebook editing tools, and it returns a boolean for success. This is specific and actionable.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies when to use it (when needing to access Colab notebook editing), but does not explicitly state when not to use it or provide alternatives. With no siblings, this is acceptable but could be more precise.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.
1 tool update
v1.0.1- First observed
open_colab_browser_connection
TDQS
Only one tool exists, so no ambiguity between tools, but the dimension assesses whether tools can be told apart; with one tool there is no need for disambiguation, but it cannot be 'clearly distinct' from others since there are none.
With a single tool, naming consistency is not applicable; however, the name is descriptive and follows a reasonable pattern, so a neutral score is given.
A single tool seems too few for a server named 'Colab MCP', which suggests a broader purpose. The tool only handles opening a connection, leaving other expected functionalities uncovered.
The server's domain appears to be Google Colab integration, but only one tool for opening a connection is provided. Missing tools for editing, running cells, managing notebooks, etc., make the surface severely incomplete.
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Connectors
Hosted MCP server connecting claude.ai, ChatGPT and other AI apps to your own computer
Remote MCP server for The Colony — a social network for AI agents (posts, DMs, search, marketplace).
An MCP server that gives your AI access to the source code and docs of all public github repos
MCP server connecting AI agents to 100+ apps (Gmail, Slack, Notion, GitHub) via one-click OAuth.
Related MCP Servers
- AlicenseBqualityDmaintenanceLocal-first MCP server for controlling Google Colab as a development, shell, file, and training runtime, with tools for notebook editing, GPU acceleration, and file transfer.597Apache 2.0
- Apache 2.0
- AlicenseNot gradedqualityFmaintenanceEnables Claude Code to execute shell commands, Python code, and file transfers on a Google Colab T4 GPU via an MCP server, bridging the GPU gap for AI coding agents.5MIT
- AlicenseNot gradedqualityDmaintenanceMCP server that connects AI agents to Google NotebookLM, enabling natural language interaction with notebooks, including Q&A, source ingestion, and audio overview generation.4,624MIT
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/aerovfx/MCP_Colab'
If you have feedback or need assistance with the MCP directory API, please join our Discord server