← Back to all workshops
Agentic Engineering with Claude Code
Replace ad-hoc 'vibe coding' with disciplined, repeatable CLI agent workflows. Learn context engineering, custom skills, MCP integrations, and test-driven AI collaboration.
Session Overview
Handing an AI agent raw access to an enterprise repository without structure leads to subtle bugs and architectural drift. Treated as a disciplined engineering partner, however, CLI coding agents like Claude Code drastically accelerate complex refactoring, debugging, and feature delivery.
Key Topics Covered
- Context Engineering & Project Briefings: Crafting layered
CLAUDE.mdfiles that enforce your team’s architectural boundaries, build commands, and coding conventions. - Disciplined Agent Loops: Structuring Plan-Act-Verify workflows where the agent writes tests, executes your Maven/Gradle suite, and self-corrects before human code review.
- Custom Skills, Hooks & MCP: Automating repetitive team workflows with custom commands, pre-commit verification hooks, and local MCP servers.
- Multi-Agent & Subagent Patterns: Delegating parallel investigation, code review, and documentation tasks without polluting your primary context window.