promptdojo_

Project workspaces, not product names — step 1 of 1

Project workspaces, not product names

By the time you read this, at least one of the tools in this lesson has probably rebranded, been acquired, or spawned a competitor with a better logo. That's fine. Chapter zero introduced the category — AI project workspaces and agentic coding tools — and the category is the part worth learning. Brand names change; the shape doesn't.

Claude Code, Codex, Cursor, Aider, OpenCode, and whatever launches next week all wrap the same basic layers around a model:

  • project files
  • tool access
  • system instructions
  • conversation history
  • diffs
  • commands
  • tests
  • permission boundaries
  • recovery checkpoints

That harness is why the same model feels different in a chat tab than in a project workspace. In chat, the model gives you an answer. In a workspace, it can inspect files, propose edits, run checks, and leave behind a diff you can review.

So the durable skill isn't memorizing one product. It's reading the harness:

  • What can the agent see?
  • What can it change?
  • What tools can it call?
  • What proof does it produce?
  • Where can I undo?
  • What should I never let it do automatically?

Answer those six questions and switching tools becomes annoying instead of existential. You understand the category.