The Developer's AI Toolkit

More Tools Than You Need (and How to Choose)

The AI tooling landscape for developers has exploded. GitHub Copilot, Cursor, Claude, Gemini Code Assist, Tabnine, Amazon Q Developer — each one promises to transform how you write software. The good news is that you do not need all of them. The better news is that once you understand the one key distinction between them, choosing becomes straightforward.

That distinction is: inline completion versus codebase-aware assistance. Every tool in this space lands somewhere on that spectrum, and where it lands determines what it is actually good at.

Inline Completion Tools

GitHub Copilot is the most widely used AI coding tool and the clearest example of inline completion done well. As you type, it suggests the next line, the next block, or sometimes the entire function. It integrates as a plugin into VS Code, Visual Studio, JetBrains, Neovim, Xcode, and Eclipse — meaning it works in whatever editor you already use. Its strengths are speed and breadth: it is genuinely fast, it handles boilerplate and repetitive code well, and it writes documentation strings faster than any human. Its limitation is that it does not have deep awareness of your overall codebase — it sees what is in the current file and recent context, not how your whole project is structured.

Gemini Code Assist integrates into VS Code and JetBrains with a similar inline approach, backed by Google's models. Amazon Q Developer (formerly CodeWhisperer) follows the same pattern but is optimised for AWS workloads — excellent if you work heavily in AWS infrastructure, more limited outside it. Tabnine differentiates on privacy: it can run entirely locally, train on your team's private codebase, and never send code to the cloud. The tradeoff is that its raw suggestion quality is lower than Copilot. If your organisation has strict data residency requirements, Tabnine is often the answer.

Codebase-Aware Tools

Cursor takes a different approach. Rather than adding AI as a plugin to an existing editor, Cursor is a fork of VS Code rebuilt entirely around AI. Its codebase indexing means it understands your functions, types, imports, and patterns across every file in your project — not just the one you have open. Its Agent Mode can reason across multiple files simultaneously, run terminal commands, fix errors it encounters, and iterate until a task is done. This makes it significantly more capable for complex multi-file work, at the cost of a steeper learning curve and the need to actually migrate your editor setup.

Claude Code is Anthropic's terminal-native coding agent. Rather than an IDE, it runs in your terminal and can read files, write files, run commands, and execute multi-step tasks with full system access. It does not provide inline completions but excels at the kinds of tasks that require coordinating across many files: large refactors, architectural changes, or building a new feature from a spec. Many developers run Cursor for moment-to-moment editing and Claude Code in the terminal for larger coordinated tasks.

Which One Should You Start With?

If you use VS Code and want the least friction: GitHub Copilot. It installs in two minutes and improves your existing workflow immediately. If you are ready to invest in a workflow change for significantly deeper AI integration: Cursor. If you already use JetBrains IDEs: Claude Code in terminal + Copilot in IDE is the current high-quality combination.

The most important thing to avoid is tool paralysis — spending more time evaluating tools than using them. Pick one and use it seriously for two weeks. The productivity gains come from depth of usage, not breadth of installation.

One Tool Is Not the Answer

As you progress through this course, you will find that different lessons naturally involve different tools. Pair programming and inline editing? Copilot or Cursor. Complex multi-file reasoning? Cursor's Agent Mode or Claude Code. Code review and debugging? Any chat interface — Claude, ChatGPT, Copilot Chat — works well. Building AI features into your app? That is the API, not an IDE plugin. Knowing which category a task falls into is the meta-skill this course is designed to build.