Catalog My learning Blog
← All courses
AI Engineering · Beginner → Advanced

AI-Assisted Coding Course: Agentic Development in 30 Days

Learn to build real software with AI coding agents such as Claude Code, Cursor, GitHub Copilot and Codex, and to keep the quality, security and understanding that vibe coding gives up. You start with how coding agents work, why they forget, how to choose a tool and what the productivity research actually shows. Then you learn context engineering: AGENTS.md and CLAUDE.md files, writing specs instead of wishes, plan mode and spec-driven development. The middle of the course covers daily workflows (test-driven development, debugging, refactoring legacy code, reviewing AI diffs, parallel agents with git worktrees) and customizing an agent with skills, hooks, subagents, MCP servers and permissions. The last phases cover AI code security, slopsquatting, prompt injection against coding agents, quality gates, headless agents in CI, AI code review, team adoption and an end-to-end capstone. Every lesson has a worked example, a diagram, real-world cases and review questions.

Step 1 of 6 in the AI engineer path. Up next: Build AI Agents with MCP: 30-Day Course.