PromptJack
Interactive two-panel simulator showing how prompt injection attacks hijack AI agents across four real-world attack scenarios — step by step.
PromptJack makes OWASP Agentic Top 10 #1 (Prompt Injection) visceral: a dark-themed two-panel layout shows what the agent intended to do on the left while the attacker's injected instruction hijacks it on the right. Four scenarios — indirect injection via fetched URL, tool description poisoning, memory persistence attacks, and multi-agent hop privilege escalation — each play out step by step with click-through simulation. Backed by real CVEs and OWASP 2026 data showing 73% of live AI rollouts are vulnerable. Mitigations drawer included.
The build log
Follow the work.
New essays and prototypes, in your feed reader. Email subscriptions are paused while I connect a working delivery service. No email is collected here.
Follow via RSS