AI for developers
Coding with AI, expert level: multi-agent, specifications and MCP server
Orchestrate several agents, drive code with specifications and create your own MCP server.
- Duration
- 7 h · 1 day
- Level
- Advanced
- Format
- On site · Remote · Blended
- Max. group size
- 12 participants
Overview
This advanced day is for developers who already practice assisted development. It covers orchestrating several agents that collaborate on a complex task, specification-driven development to get consistent, maintainable code, then creating, configuring and securing an MCP server.
Who this program is for
- Experienced developers
- Tech leads
- Architects
Prerequisites
- Practical command of at least one programming language
- Daily use of Git and a modern IDE
- Habit of AI-assisted development
Learning objectives
- Design multi-agent systems that collaborate
- Write technical specifications an agent can use
- Create, configure and secure an MCP server
Detailed curriculum
2 modules · 7 h · 1 day
Day 1
Multi-agent orchestration and specification-driven development
30% theory · 70% practice
- Roles and delegation between agents
- Orchestration patterns
- Quality and cost control
- Writing a usable technical specification
- From specification document to validated code
🛠 Hands-on workshops
Agent chain1.5 h
Make several agents collaborate on a feature.
✅ Deliverable : A multi-agent workflow
Create your MCP server
30% theory · 70% practice
- Reminders on the MCP architecture
- Designing the tools to expose
- Implementing with an SDK
- Authentication and permissions
- Testing and deployment
🛠 Hands-on workshops
Team MCP server2 h
Create an MCP server useful to your team.
✅ Deliverable : A secured MCP server
Tools and technologies
Claude CodeGitMCPVisual 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
- A modern IDE, an advanced code assistant and Git
Resources provided
- Course materials and repository of corrected exercises
Career outcomes
- A documented multi-agent workflow
- A reusable specification template
- A working MCP server