AI for developers
MCP: give your AI models tools and context
Understand the Model Context Protocol, create MCP servers and deploy them safely.
- Duration
- 7 h · 1 day
- Level
- Advanced
- Format
- On site · Remote · Blended
- Max. group size
- 12 participants
Overview
This day presents the Model Context Protocol (MCP), a standard that lets a model access tools and data in a uniform way. Participants understand its architecture, create a server exposing resources and tools with an SDK, then tackle advanced topics: standardized prompts, sampling, scopes, logging, testing, authentication and security. The goal is to be able to design modular architectures built on MCP.
Who this program is for
- Developers and architects
- Platform engineers and DevOps
- Teams equipping AI assistants
Prerequisites
- Comfort with Python (or a language supported by an MCP SDK)
- Basic understanding of LLMs and APIs
Learning objectives
- Explain the value and architecture of MCP
- Distinguish hosts, clients and servers
- Expose resources, tools and prompts
- Connect a model to tools and data
- Inspect, test and debug a server
- Apply authentication and security best practices
Detailed curriculum
2 modules · 7 h · 1 day
Day 1
Discovering MCP
30% theory · 70% practice
- The problem the protocol solves
- Hosts, clients, servers and data sources
- Protocol layers and connection lifecycle
- Resources, tools and prompts
- Available SDKs
🛠 Hands-on workshops
First MCP server1.5 h
Create a server that exposes a tool and a resource.
✅ Deliverable : A working server
Advanced concepts and production
30% theory · 70% practice
- Standardized prompts and sampling
- Server scope
- Logging and error handling
- Inspection, testing and debugging
- Authentication, access control and sensitive data
🛠 Hands-on workshops
Secure your server1.5 h
Add authentication and controls to a server.
✅ Deliverable : A secured server
Tools and technologies
Claude CodeMCPPythonVisual Studio Code
Teaching methods
- Alternating short lectures and hands-on work
- Live demonstrations by the trainer
- Capstone project built throughout the modules
- Group corrections and discussion of best practices
Assessment
- Hands-on exercises corrected in each module
- Mini-project completed at the end of the course
- Knowledge-check quiz
Required equipment
- A computer with an internet connection
- Python 3 or Node.js, a code editor and an MCP-compatible client
Resources provided
- Course materials and repository of corrected exercises
Career outcomes
- A working MCP server
- A reference architecture
- A security checklist