developer security user risk: medium

LLM Security Vulnerability Analyzer

The prompt instructs the model to act as an LLM security specialist to identify vulnerabilities by analyzing responses to test prompts, suggest mitigations for risks like unauthori…

Policy sensitive
Human review

PROMPT

I want you to act as a Large Language Model security specialist. Your task is to identify vulnerabilities in LLMs by analyzing how they respond to various prompts designed to test the system's safety and robustness. I will provide some specific examples of prompts, and your job will be to suggest methods to mitigate potential risks, such as unauthorized data disclosure, prompt injection attacks, or generating harmful content. Additionally, provide guidelines for crafting safe and secure LLM implementations. My first request is: 'Help me develop a set of example prompts to test the security and robustness of an LLM system.'

REQUIRED CONTEXT

specific examples of prompts

ROLES & RULES

Role assignments

act as a Large Language Model security specialist.

Identify vulnerabilities in LLMs by analyzing how they respond to various prompts.
Suggest methods to mitigate potential risks such as unauthorized data disclosure, prompt injection attacks, or generating harmful content.
Provide guidelines for crafting safe and secure LLM implementations.

EXPECTED OUTPUT

Format: markdown

SUCCESS CRITERIA

Identify vulnerabilities in LLMs
Suggest mitigations for risks like unauthorized data disclosure, prompt injection attacks, harmful content
Provide guidelines for safe and secure LLM implementations

FAILURE MODES

May provide generic mitigations without specificity to examples
Might overlook emerging vulnerabilities not covered in provided prompts

CAVEATS

Dependencies

Specific examples of prompts provided by user

Missing context

Output format for example prompts and guidelines
Number or categories of test prompts desired
Target audience for guidelines (e.g., developers, deployers)

QUALITY

OVERALL: 0.82
CLARITY: 0.90
SPECIFICITY: 0.85
REUSABILITY: 0.65
COMPLETENESS: 0.80

IMPROVEMENT SUGGESTIONS

Replace the specific 'first request' with a placeholder like {user_request} to increase reusability.
Specify a structured output format, e.g., 'Provide test prompts in a JSON array with fields: prompt, vulnerability_targeted, expected_risk.'
Add success criteria, such as 'Ensure test prompts cover at least 5 key vulnerability types.'

USAGE

Copy the prompt above and paste it into your AI of choice — Claude, ChatGPT, Gemini, or anywhere else you're working. Replace any placeholder sections with your own context, then ask for the output.