AI Prompter & Autocomplete
Write structured prompts naming inputs, outputs, and edge cases โ plus a Codex-style comment-driven completion challenge โ graded against a simulated AI response.
AI Prompting, Security & Agentic Coding Simulator
Learn to work WITH AI coding tools. Write prompts that actually get useful results, harden a chatbot against jailbreak attacks, supervise an autonomous coding agent's proposed actions, and audit AI-generated code for bugs and hard-coded secrets.
Grade 9โ12 students, aspiring developers, and anyone learning to collaborate with AI coding tools responsibly.
Write structured prompts naming inputs, outputs, and edge cases โ plus a Codex-style comment-driven completion challenge โ graded against a simulated AI response.
Write a system prompt, then watch simulated jailbreak attacks try override claims, roleplay framing, and indirect disclosure โ harden it against all of them.
Review an AI coding agent's proposed steps as the human overseer โ approve safe, on-task actions and reject destructive commands, scope creep, and hard-coded secrets.
Fix broken tags and CSS selectors in AI-generated code, and audit for security issues โ plaintext passwords and hard-coded API keys hiding in comments.
40 quiz questions across prompt engineering, AI security, agentic oversight, and code auditing โ each with a full explanation, so wrong answers teach too.
Earn Prompt Architect, Jailbreak Defender, Agent Director, and Security Auditor badges, climb the leaderboard, and unlock a completion certificate at 100%.
Completely self-contained sandbox โ no internet required when run locally in the classroom.
Educational simulation only โ all AI responses, attacks, and agent actions are locally simulated. No real API calls are ever made.
Every session ends with a quick 15โ20 min real-world task that turns the simulator into a habit โ research and observation only, no accounts or spending.
Write one vague prompt and one specific prompt for the same small task, and compare (or predict) how the results would differ.
๐ Safety-first โ never attempt real jailbreaks against production AI systems; this lab is simulated locally for exactly that reason.
VibeCoder & Agentic Studio is part of our hands-on Life Skills workshops.
No payment required to apply. We'll contact you to confirm schedule options and availability.