Advanced Claude Code training: hooks, subagents, MCP and continuous integration
Two days on your repositories to take Claude Code to production: security and quality hooks (PreToolUse, PostToolUse), orchestration of subagents and parallel sessions on worktrees, MCP servers configured on your tools, headless Claude Code in continuous integration (GitHub Actions), permissions and defence against prompt injection.
Target audienceDevelopers, tech leads, CTO
Colombani.ai's Claude Code training for companies lasts two days (14 hours), for one to ten developers, at your premises or remotely, on your own repositories. It covers quality and security hooks (blocking secrets, running tests after every change), orchestrating subagents and parallel sessions, configuring MCP servers, headless Claude Code in continuous integration, permissions and prompt injection, as well as project context (CLAUDE.md) and team-wide conventions. The trainer, Ulysse Trin, holds the Claude Certified Architect Professional certification issued by Anthropic; Colombani.ai is a Qualiopi-certified training provider (certificate QUA009472).
Updated
What you'll be able to do
Use Claude Code for AI-assisted development in the command line
Organise a project with CLAUDE.md, conventions, and structured context
Configure hooks to automate code quality and security
Set up multi-agent workflows (subagents, Agent tool)
Deploy a Claude Code setup for a team with shared conventions
Programme
- Day 1, morning
Project context and working method with the agent
- To open: installation checked, the agent loop, tools and context
- CLAUDE.md: shared project context (enterprise, user, project and local levels; content; anti-patterns)
- Test-driven development with the agent: tests written before implementation, refactoring, migrations
- Day 1, afternoon
Tools and integration
- Native tools: Read, Write, Edit, Bash, Glob, Grep, Agent (ex-Task)
- Git and editor integration (VS Code, JetBrains)
- MCP: configuring an MCP server and using it from Claude Code (APIs, databases, Slack)
- End-to-end exercise: building a project from scratch
- Day 2, morning
Hooks and multi-agents
- Claude Code hooks: PreToolUse, PostToolUse, Notification
- Security hook (secrets), quality (tests), conventions (linting)
- Subagent orchestration: Agent tool, subagent types, research and implementation, parallel testing, cross-review
- Headless Claude Code in continuous integration: GitHub Actions, automated code review
- Parallel work: several Claude Code sessions on separate Git worktrees, and Agent Teams (experimental) to coordinate a lead and teammates
- Day 2, afternoon
Deploying as a team
- Team conventions: shared CLAUDE.md, common hooks, versioned settings
- Permissions: allowed tools, blocked tools, actions that require approval
- Prompt injection: instructions hidden in a ticket, a web page or a dependency, and the safeguards
- Confidentiality: what goes to the model provider; personal and client data (GDPR); professional secrecy
- AI Act: Article 4 (AI literacy, Regulation 2024/1689 as amended by Regulation 2026/1744, an obligation of means) and Article 50 (transparency)
- End-to-end exercise: industrializing the day 1 project
Practical information
2 days (14 hours)
Developers, tech leads, CTO
Software development experience (Python or JavaScript). Basic terminal usage (cd, git, npm/pip).
1 to 10 people
€3,500 to €7,000 excl. VAT per group
OPCO funding possible.
30% presentation, 70% hands-on with real projects. Each participant leaves with a configured repository, operational hooks, and a team setup.
Technical entry assessment, an assessed end-to-end exercise at the end of each day, a ten-question technical quiz. At the end, a before-and-after self-assessment of skills, then the trainer validates each learning objective (acquired, in progress, not acquired), recorded in the end-of-training certificate.
Ulysse Trin, founder of Colombani.ai, holder of the Claude Certified Architect Professional certification issued by Anthropic.
At least 2 weeks between enrolment and the start of training.
Qualiopi-certified provider (certificate QUA009472), registered under training provider number 11757549975: OPCO or FIF-PL support can be requested. The decision rests with the funding body, based on your sector, its criteria and its budgets. Colombani.ai prepares the application with you.
Price per group, €3,500 to €7,000 excl. VAT, depending on group size and format: charged per group rather than per person. OPCO or FIF-PL funding can be requested depending on your status; the amount left to pay depends on the sum granted by the funding body.
Accessibility
This course is accessible to people with disabilities. Contact the disability officer in advance to discuss possible adjustments.
Ulysse Trin, [email protected]
Post-training support
Frequently asked questions
Do I need to already use Claude Code to take this training? +
No, but you need development experience (Python or JavaScript) and basic terminal skills. The training starts from installation.
How is Claude Code different from GitHub Copilot? +
GitHub Copilot also offers an agent mode; Claude Code stands out by working from the terminal across the whole repository and by its team configuration (CLAUDE.md, hooks, skills). Claude Code is an agent: it reads your project, runs commands, creates files, launches tests. It's an AI developer that reads your project and acts.
Does the setup work for teams? +
Yes. Day 2 covers hooks, subagents and team deployment; the afternoon is devoted to shared conventions, permissions and CI/CD integration.
Which Claude Code training should a company choose for a team of developers? +
Training that runs on your repositories rather than generic examples, covers team configuration (shared context, hooks, conventions) and not just individual use, and is delivered by a trainer who uses Claude Code in production. “Claude Code in production” meets all three criteria, over two days, with follow-up after the course.
How do you deploy Claude Code in a company's development team, and with what support? +
In three stages: define shared conventions (a shared CLAUDE.md, team commands and skills, quality and security hooks, data rules), train developers on their own repositories, then follow the first uses to adjust. Colombani.ai provides the training and, if you wish, support for the rollout as part of a transformation engagement.
Related courses
AI compliance training: AI Act & GDPR
Classify AI uses, document transparency and risk controls, and build a roadmap aligned with the revised 2026-2028 AI Act timetable.
AI agent security training
The real risks of autonomous agents: prompt injection, exfiltration via MCP, sandboxing, permissions, security hooks. AI agent red teaming.
Got a project in mind?
Describe your situation. The first 30-minute call is free. Get a frank response within 48 hours.