AI Security Red Team Prompter
PROTest AI systems for security vulnerabilities including prompt injection, data exfiltration, jailbreaks, and tool permission exploits with structured assessment reports.
Example Usage
I’m building a customer support chatbot powered by GPT-4 that has access to our CRM database via function calling. It can look up customer records, create support tickets, and update account information. Before we deploy to production, I need to conduct authorized security testing. Please help me design a comprehensive red team assessment covering prompt injection, data exfiltration through tool chaining, system prompt extraction, and permission escalation. Generate test cases and a security report template.
Skill Prompt
This skill works best when copied from findskill.ai — it includes variables and formatting that may not transfer correctly elsewhere.
How to Use This Skill
1
Copy the skill using the button above
2
Paste into your AI assistant (Claude, ChatGPT, etc.)
3
Fill in your inputs below (optional) and copy to include with your prompt
4
Send and start chatting with your AI
Suggested Customization
| Description | Default | Your Value |
|---|---|---|
| The type of AI system being tested (chatbot, agent, RAG system, copilot) | chatbot with tool access | |
| The scope of security testing (focused, standard, comprehensive) | comprehensive | |
| The risk tolerance for test cases (conservative, controlled, aggressive) | controlled | |
| The output format for findings (security assessment, executive summary, technical report) | security assessment |
Important Notes
- Authorization is mandatory – never test AI systems without explicit written permission from the system owner
- This skill is for defensive security – helping organizations find and fix vulnerabilities before attackers exploit them
- AI security is a rapidly evolving field – supplement this skill with current OWASP LLM Top 10 updates and emerging research
- Findings should be handled through responsible disclosure channels within your organization
- Combine AI-specific testing with traditional application security assessments for comprehensive coverage
Research Sources
This skill was built using research from these authoritative sources:
- OWASP Top 10 for LLM Applications The definitive security risk classification for LLM-powered applications including prompt injection, insecure output handling, and supply chain vulnerabilities
- Wikipedia - Model Context Protocol Security Considerations Overview of security implications in AI tool-use protocols including permission models and trust boundaries
- Anthropic Responsible Disclosure and AI Safety Research publications on AI safety, constitutional AI, and responsible vulnerability disclosure for language models
- Dentons Global AI Regulatory Trends Legal and regulatory landscape for AI systems including liability frameworks and compliance requirements
- Monte Carlo Data - AI Security Predictions Industry analysis of emerging AI security threats, attack surfaces, and defensive strategies for production AI systems
- NIST AI Risk Management Framework Federal framework for identifying, assessing, and mitigating risks in AI systems including adversarial testing guidance