Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Course Outline
MODULE 1: Claude Code vs. Traditional AI Coding Tools
- Why “coding assistant” is the wrong mental model for Claude Code
- Architecture comparison: terminal-native agent vs. IDE inline completion
- The agentic cycle: read → plan → execute → verify → report
- What Claude Code can do autonomously; where human sign-off is required
- Comparing strategies: agentic delegation vs. inline completion vs. chat-based iteration
- Practical decision matrix: which approach to use and when
MODULE 2: Installation, Configuration & the Permissions Model
- Installing Claude Code and configuring API credentials
- Trust and permissions model: file access, shell commands, network calls
- Auto-accept mode vs. interactive approval (speed vs. safety trade-offs)
- Reading and understanding execution plans before running
- Interrupting, redirecting, and resuming tasks without losing state
- Project-level settings in .claude/settings.json
MODULE 3: CLAUDE.md – Persistent Project Memory
- Why LLMs need explicit context (stateless sessions)
- CLAUDE.md as a living project knowledge file
- Architecture, conventions, and tech stack documentation
- Global vs. project vs. directory-level CLAUDE.md
- Encoding security constraints (what must never be done autonomously)
- Maintaining CLAUDE.md without creating noise
- Hands-on: create a CLAUDE.md for a real project
MODULE 4: Writing Delegations That Work
- Anatomy of a well-formed task: context, constraints, acceptance criteria, output format
- Why simple prompts fail and how to structure proper delegations
- Staged delegation: breaking complex work into checkpoints
- Delegating exploration before implementation
- Diagnosing and preventing failure modes
- Practice: delegate progressively complex tasks
MODULE 5: Exploring & Documenting Unfamiliar Codebases
- Fast onboarding to unknown repositories
- Generating architecture maps and module summaries
- Identifying undocumented behaviour and technical debt
- Creating living documentation updated with code changes
- Practice: work with an unfamiliar open-source project
MODULE 6: Code Review, Refactoring & Quality Assurance
- Delegating structured code reviews (security, design, performance)
- Multi-file refactoring with traceability
- Human-in-the-loop validation during refactoring
- Evaluating AI-generated code quality
- Common anti-patterns (blind acceptance, lack of verification)
MODULE 7: Skills, Agents & Automation Hooks
- Creating reusable skills (shared command workflows)
- Building sub-agents for specific tasks
- Pre-tool and post-tool hooks for automation
- Sharing workflows across repositories
- Orchestrating multi-step workflows using agents
MODULE 8: Model Context Protocol (MCP): Connecting Claude Code to Your Tools
- MCP fundamentals: client–server model, JSON-RPC 2.0
- Tools, resources, and prompt primitives
- Connecting MCP servers (GitHub, filesystem, PostgreSQL, Slack, Linear)
- Building a custom MCP server in Python
- Security boundaries and access control
- Practice: connect at least one MCP server
MODULE 9: Parallel Development: Worktrees & Subagents
- Limits of sequential single-agent workflows
- Using Git worktrees for parallel development
- Running multiple Claude Code sessions simultaneously
- Coordinating subagents for parallel tasks
- Managing merge conflicts and coordination overhead
MODULE 10: Designing with Claude Code
- Using natural language for UI and design tasks
- Generating and iterating front-end components (HTML, CSS, behaviour)
- Managing and evolving design systems
- Integrating design tools via MCP (Figma, Storybook)
- Evaluating and refining AI-generated designs
MODULE 11: Team Standards, Security & Governance
- Shared CLAUDE.md templates across repositories
- Defining hard constraints for autonomous execution
- Data privacy and IP protection
- Audit trails and activity tracking
- Enterprise considerations: compliance, filtering, data residency
MODULE 12: Designing Your AI Development Workflow
- Mapping current development pain points
- Designing personal and team workflows
- Combining CLAUDE.md, MCP, agents, and parallel execution
- Measuring effectiveness and productivity gains
- Creating actionable improvement plans
Requirements
Prerequisites
- Solid software development experience in at least one programming language
- Comfortable working in a terminal or command-line environment
- Basic familiarity with Git (commits, branches, pull requests)
- Prior exposure to an AI coding tool (such as GitHub Copilot, Cursor, or ChatGPT for coding) is helpful, but not required
- Claude code Licence!
Target Audience (Who This Course Is For)
- Software engineers who already use AI coding tools and want to move towards agentic workflows
- Technical leads and senior developers evaluating Claude Code for team adoption
- Engineers transitioning from IDE-embedded assistants to agent-first development approaches
- DevOps and platform engineers interested in integrating AI into automated pipelines
- Engineering managers planning a team-wide adoption of AI-assisted development and defining governance models
14 Hours
Testimonials (2)
Learning how to prompt Claude and use it to digest all of the data I have available.
Mike Hartleroad - Furniture Row
Course - Claude AI for Data Analysis and Business Intelligence
how to engage with the Office environment and set up repetitive tasks
