> ## Documentation Index
> Fetch the complete documentation index at: https://docs.enokilabs.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Concepts

> Targets, assessments, findings, and how grading works.

The vocabulary: a **target** (your agent, registered with a connection
config), an **assessment run** (adaptive attacks against it), **findings**
(validated failures with severity and a lifecycle), and the **judge** that
grades every exchange fail-closed.