Verifying AI Output
Unscored tasks with hints and instant feedback. Practice results never become credentials.
Short lesson
Name each agent by role and task, isolate them in worktrees, hand off through files, review every diff, never paste secrets. Tool notes (2026-10): Claude Code (CLAUDE.md, agents folder), Codex CLI and others (AGENTS.md), Gemini CLI (GEMINI.md), Cursor, Copilot — verify names in each tool's docs.
An AI answered “How do I list hidden files on macOS?”. Click every incorrect line.
Sign up to practiseThe AI wrote this to compute the average order value from a list of orders. Click every line with a bug.
Sign up to practiseSource data: 2,140 survey responses; 61% use AI weekly; top use is drafting emails (44%). Click every claim the AI made that is wrong or unsupported by the data.
Sign up to practiseYou let an agent refactor 30 files. What gives the strongest assurance before merging?
Sign up to practise