Academy
What is prompt injection?
Glossary · AI agent and machine identity security
Short answer
Prompt injection is an attack in which text that an AI agent reads, such as a web page or ticket, contains instructions that change what the agent does.
Why it matters
A successful injection can only use the access the agent already has. That makes the agent's reach the limit of the damage.
What it looks like
A support ticket with hidden instructions makes an agent with database access reveal private data.
In nerl0
nerl0 does not inspect prompts. It shows what an agent can reach, which is the limit of what a successful injection could touch. See Incident Tracker.