AI Red Team Lab

AI Red Team Lab Single-turn

Connected to Backend Engine

Target LLM

empty

Judge LLM

empty

Attacker LLM

empty

⚡ Live Red Team Execution

Idle

Current Test Item

Waiting to start...

Remaining Queue

-

Execution Log

Configuration Setup

⚠ Your API key can be stolen if your browser or endpoint is unsafe.

Test Name (used for Report ID)

Prompts Per Test Item

Judge Rubric

Prompt Language

Multi-Turn Sliding Window

Target LLM

F5 Guardrail

ML Threshold

Judge LLM

Attacker LLM

Test Items 0 selected

Custom Test Intent

Reports & Raw Data

Reuse Existing Dataset

Upload Dataset (CSV)

CSV format: prompt
Max: 200 prompts

Block Detection Signals Guardrail Detection

Define keywords that indicate a prompt was blocked by a guardrail or ML safety engine.

Signal 1

Signal 2

Signal 3

Signal 4

Signal 5

Target LLM

Judge LLM

Attacker LLM

⚡ Live Red Team Execution

Current Test Item

Remaining Queue

Execution Log

Test Name (used for Report ID)

Prompts Per Test Item

Judge Rubric

Judge LLM

Attacker LLM

Report Details