Development Tools

AI coding tools are development environments where an AI model doesn’t just autocomplete your code — it reads your codebase, plans changes, edits files, runs commands, and iterates on failures. The category spans several shapes: agentic CLIs like Claude Code and Codex that work in your terminal, AI-native IDEs like Cursor and Windsurf that rebuild the editor around the model, and cloud platforms like Replit that pair an agent with hosting so an idea becomes a deployed app in one place.

This module matters for two reasons. First, these tools have changed how software gets built: developers who direct an agent well ship meaningfully faster than those typing every line, and knowing when to delegate versus when to intervene is now a core skill. Second, as an AI engineer, these tools are your best case study — they’re production agent systems whose architecture (context gathering, tool use, planning, self-correction, sandboxing) embodies everything this course teaches about building agents.

In practice, you’ll pick a primary tool and learn to drive it well: writing clear task descriptions, maintaining project context files (like CLAUDE.md or Cursor rules) that teach the agent your conventions, reviewing diffs rather than rubber-stamping them, and using MCP servers to connect agents to your databases and services. The following topics walk through each major tool so you can choose based on your workflow — terminal, IDE, or browser — rather than hype.