Introduction to MCP Functions

MCP Functions is the first AI-powered platform for creating Model Context Protocol (MCP) tools. Transform your business with intelligent automation - no code, no deployment, no infrastructure needed.

What is Model Context Protocol (MCP)?

To understand MCP Functions, you first need to understand what the Model Context Protocol (MCP) is and why it matters. Think of MCP as a universal language that allows AI assistants to communicate with external tools and services.

Imagine you have an AI assistant like ChatGPT or Claude. By default, these AI models can only work with the information they were trained on. They can't access real-time data, interact with your company's databases, send emails, create calendar events, or perform any actions in the real world. This is where MCP comes in.

The Model Context Protocol is a standardized communication system that enables AI assistants to:

• Discover available tools - The AI can ask "what tools do you have?" and get a list of everything available

• Understand what each tool does - Each tool provides a description of its purpose and what parameters it needs

• Invoke tools with the right data - The AI can call tools with specific inputs and receive structured responses

• Handle errors gracefully - If something goes wrong, the protocol ensures the AI understands what happened

Why MCP Exists

Before MCP, if you wanted to connect an AI assistant to your business tools, you had to build custom integrations for each AI platform. This meant:

• Creating different code for ChatGPT, Claude, and other AI assistants

• Maintaining multiple integrations that could break when AI platforms updated

• Building authentication, error handling, and security from scratch each time

• Spending weeks or months on integration work instead of building actual business value

MCP solves this by providing a single, standardized protocol that any AI assistant can use. Once you build tools using MCP, they work with all MCP-compatible AI assistants automatically. This means you write your tool once, and it works everywhere.

How MCP Works in Practice

Here's a simple example of how MCP works:

• You create a tool - For example, a tool that fetches weather data for any city

• The tool is registered with an MCP server - The server knows about your tool and can describe it to AI assistants

• An AI assistant connects to your MCP server - The AI asks "what tools do you have?"

• The server responds with tool descriptions - "I have a weather tool that takes a city name and returns current weather"

• The AI can now use your tool - When a user asks "what's the weather in New York?", the AI calls your tool with "New York" as the parameter

• Your tool executes and returns data - The weather data flows back to the AI, which presents it to the user

This entire process happens automatically. The AI doesn't need to know how your tool works internally - it just needs to know what inputs it takes and what outputs it provides. This is the power of standardization.

What is MCP Functions?

MCP Functions is a platform that makes creating and managing MCP tools incredibly simple. Traditionally, building MCP tools required:

• Setting up servers to host your MCP tools

• Writing code to implement the MCP protocol

• Configuring databases, authentication, and security

• Building deployment pipelines and monitoring systems

• Maintaining infrastructure 24/7

MCP Functions eliminates all of this complexity. Instead of spending weeks setting up infrastructure, you can create powerful MCP tools in minutes using natural language. Our AI understands what you want to build and generates the code automatically. Your tools are deployed instantly and run securely in our cloud infrastructure.

The Problem MCP Functions Solves

Let's say you want to create a tool that helps your AI assistant access your company's customer database. Without MCP Functions, you would need to:

• Set up a server (AWS, Google Cloud, or Azure) - 1-2 days

• Configure the server with Node.js, Python, or another runtime - 1 day

• Write code to implement the MCP protocol - 2-3 days

• Write code to connect to your database - 1 day

• Set up authentication and security - 2-3 days

• Create deployment scripts - 1 day

• Set up monitoring and logging - 1-2 days

• Test everything thoroughly - 2-3 days

Total time: 2-3 weeks of developer time, plus ongoing maintenance.

With MCP Functions, you simply describe what you want: "Create a tool that queries our customer database and returns customer information by email address." Our AI generates the code, deploys it instantly, and your tool is ready to use. This entire process takes minutes, not weeks.

Key Features of MCP Functions

The platform includes:

• Streaming and agent-based AI code generation — Generate code in real time (streaming) or use the LangGraph agent with optional web search; the AI can also suggest tool name, description, and parameters.

• Tool flow diagrams — Automatically generate business flow diagrams from your tool definition.

• Tool config — Reference workspace secrets and OAuth2 in tool configuration; both are resolved securely at execution time.

• Full MCP support — initialize, tools/list, tools/call, resources/list, resources/read, prompts/list, prompts/get, logging/setLevel.

• Flexible server auth — API key (Bearer header or query), optional custom headers; require one or more methods per MCP server.

• Tool versions — Tools are versioned; publish when ready and view or roll back to previous versions.

• Dashboard — Organization statistics, invocation charts, activity feed, and country-level invocations.

• Sandbox — VM-based execution with configurable timeouts, HTTP limits, and blocked patterns for security.

Zero Infrastructure Required

You don't need to provision servers, configure databases, set up load balancers, or manage any infrastructure. Everything runs on our secure, scalable cloud platform. When you create a tool, it's automatically deployed and available immediately. As your usage grows, the platform scales automatically without any configuration on your part.

AI-Powered Tool Creation

Instead of writing code from scratch, you describe what you want in plain English. Our AI understands your requirements and generates production-ready code automatically. For example, you might say "Create a tool that sends a Slack message to a channel" and our AI will generate the complete code, including error handling, parameter validation, and proper MCP protocol implementation.

Instant Deployment

There's no build process, no deployment pipeline, and no waiting. The moment you create or update a tool, it's live and ready to use. This means you can iterate quickly, test ideas immediately, and respond to business needs in real-time.

Secure Sandbox Execution

Every tool runs in an isolated, secure environment called a sandbox. This means:

• Your tools can't access other tools' data or code

• Malicious code is contained and can't harm your system

• Each execution is completely isolated from others

• Resource limits prevent runaway processes

This security is built-in and automatic - you don't need to configure anything.

Team Collaboration

Tools are shared instantly with your team. When you create a tool, everyone in your organization can use it immediately. There's no need to coordinate deployments, share code repositories, or manage access separately. Team members can see all available tools, understand what each one does, and use them in their AI assistants right away.

Enterprise-Grade Reliability

The platform is built with enterprise needs in mind:

• High availability - 99.9% uptime guarantee with automatic failover

• Automatic scaling - Handles everything from a few requests to millions per day

• Comprehensive logging - Every tool execution is logged for debugging and auditing

• Version control - Track changes, rollback to previous versions, and maintain history

• Rate limiting - Protect your tools from abuse automatically

How MCP Functions Organizes Your Tools

MCP Functions uses a hierarchical structure to organize your tools logically. This structure helps you manage complex projects, separate different environments, and control access appropriately. Let's explore each level:

Organization The top-level container that represents your company or team. An organization contains all your workspaces, MCP servers, and tools. It's where you manage team members, billing, and organization-wide settings. Think of it as your company's account on the platform. Example: "Acme Corporation" is an organization that contains all of Acme's MCP tools and resources. ↓ Workspace An isolated environment within your organization where you group related projects and tools. Workspaces help you separate different projects, environments (like development and production), or teams. Each workspace has its own MCP servers and tools, and you can control who has access to each workspace. Example: Within "Acme Corporation", you might have workspaces like "Customer Support Tools", "Sales Automation", and "Internal Operations". ↓ MCP Server A container that groups related tools together and exposes them via the MCP protocol. An MCP server is what AI assistants connect to when they want to use your tools. You can think of it as a "toolbox" - it contains multiple tools that work together to solve a particular problem or serve a specific purpose. Example: Within the "Customer Support Tools" workspace, you might have an MCP server called "Customer Data Tools" that contains tools for looking up customer information, updating customer records, and creating support tickets. ↓ Tool An individual function that performs a specific task. Each tool has a name, description, input parameters, and code that executes when called. Tools are the building blocks - they're what actually do the work when an AI assistant needs to perform an action. Example: Within the "Customer Data Tools" MCP server, you might have individual tools like "Get Customer by Email", "Update Customer Status", and "Create Support Ticket".

Why This Structure Matters

This hierarchical organization provides several important benefits:

• Logical organization - Related tools are grouped together, making them easy to find and manage

• Access control - You can grant different team members access to different workspaces or servers

• Environment separation - Keep development, staging, and production tools separate

• Scalability - As your organization grows, this structure helps you manage hundreds or thousands of tools

• Clear ownership - It's clear which team or project each tool belongs to

What You Can Build with MCP Functions

The possibilities are nearly endless. Here are some examples of what you can create:

• Database tools - Query your databases, update records, generate reports

• API integrations - Connect to external services like Slack, Salesforce, Stripe, or any REST API

• File operations - Read, write, and process files in cloud storage

• Communication tools - Send emails, SMS messages, or Slack notifications

• Data processing - Transform data, perform calculations, generate insights

• Workflow automation - Automate complex business processes

• Custom business logic - Implement any logic specific to your business needs

And because these tools work through the MCP protocol, they're immediately available to any AI assistant that supports MCP - whether that's ChatGPT, Claude, or future AI platforms.

Getting Started

Now that you understand what MCP is and what MCP Functions does, you're ready to start building. The next steps are:

• Set up your account - Create an organization and configure your team

• Create your first workspace - Set up an isolated environment for your tools

• Build your first tool - Use our AI-powered creation to build something useful

• Connect it to an AI assistant - See your tool in action

Don't worry if this all seems new - the platform is designed to be intuitive, and our documentation will guide you through every step. Let's begin!