PromptInjection - LLM Prompt Injection Security Testing
Test LLM applications for prompt injection vulnerabilities and security risks
Tags
Updated: 2026-05-11prompt-injectionLLM-securityjailbreak-testingsecurity-assessmentpentestingvulnerability-testingAI-security
Capabilities
What this skill does
- execute direct injection tests
- execute indirect injection tests
- execute multi-stage attacks
- map application attack surface
- validate jailbreak attempts
- extract system prompts
- test guardrail bypasses
- analyze RAG systems
- generate security reports
Inputs
- written authorization
- target system URL
- system scope documentation
- attack techniques documentation
- browser automation access
- local notification endpoint
Outputs
- voice notifications
- workflow status messages
- vulnerability test results
- security assessment reports
- remediation recommendations
Requirements
- written authorization required
- browser automation environment
- localhost:8888 notification service
- authorized testing scope