Internal Tool — v1.0
Stored in memory only. Never sent anywhere except api.anthropic.com
1
Client Intake
2
Attack Prompts
3
Generate Report
Client Information
Attack Surface
Prompt Injection
Jailbreaking
System Prompt Extraction
Roleplay Exploitation
Tool Misuse
Indirect Injection
Data Exfiltration
Goal Hijacking
Privilege Escalation
Insecure Output
Engagement Preview
Client
Not set
Product
Not set
Description
Not set
Model
Not set
Tool Access
Not set
Selected Vectors
None selected
Attack Prompts
⚡

Generate attacks from
the intake form first.

→

Select an attack prompt
from the sidebar to begin testing.

Report Generator
0
Vulnerabilities
0
Passed
0
Total Tested
Report Preview
Report will appear here once generated...