prompt-plus-plus-mcp/metaprompts/advanced_thinking/adversarial.json

Advanced MCP server with 44+ metaprompt strategies including AI Core Principles, Vibe Coding Rules, and metadata-driven intelligent selection

{
  "name": "Adversarial Red-Blue Team Engine",
  "description": "Strategic adversarial thinking using red team (attack/critique) vs blue team (defend/build) methodology. Perfect for stress-testing ideas, finding vulnerabilities, and building robust solutions through opposition. Use when you need bulletproof thinking.",
  "template": "Deploy the Adversarial Red-Blue Team Engine to stress-test and fortify this prompt through strategic opposition and defense.\n\nInitial Prompt: [Insert initial prompt here]\n\n# Adversarial Framework\n\n## Red Team Operations (Attack Phase)\n\n### Vulnerability Assessment\n- **Assumption Attacks**: Challenge every unstated assumption\n- **Logic Bombs**: Find logical inconsistencies and contradictions\n- **Edge Case Exploitation**: Identify scenarios where the approach fails\n- **Resource Exhaustion**: Push the solution beyond its limits\n- **Social Engineering**: Consider human factor vulnerabilities\n\n### Strategic Deception Analysis\n- **Trojan Horse Problems**: Hidden complexities disguised as simple issues\n- **False Dilemmas**: Artificial constraints that limit thinking\n- **Motte and Bailey**: Strong positions that retreat to weaker defensible ones\n- **Strawman Variants**: Misrepresentations that weaken the approach\n- **Moving Goalposts**: Changing success criteria mid-analysis\n\n### Attack Vectors\n1. **Technical Attacks**: Find implementation weaknesses\n2. **Conceptual Attacks**: Challenge fundamental premises\n3. **Contextual Attacks**: Exploit environmental dependencies\n4. **Temporal Attacks**: Identify time-based vulnerabilities\n5. **Scale Attacks**: Break the solution at different magnitudes\n\n## Blue Team Operations (Defense Phase)\n\n### Defensive Strategies\n- **Assumption Hardening**: Make implicit assumptions explicit and defensible\n- **Redundancy Building**: Create multiple paths to success\n- **Graceful Degradation**: Ensure partial function under attack\n- **Counter-Intelligence**: Anticipate and prepare for likely attacks\n- **Adaptive Response**: Build in learning and evolution capabilities\n\n### Fortification Techniques\n- **Defense in Depth**: Multiple layers of protection\n- **Fail-Safe Design**: Safe failure modes when attacks succeed\n- **Monitoring Systems**: Early warning of attack patterns\n- **Rapid Response**: Quick adaptation to new attack vectors\n- **Counter-Attack**: Turn defensive positions into offensive advantages\n\n## Purple Team Integration (Synthesis Phase)\n\n### Collaborative War Gaming\n- **Scenario Planning**: Red team proposes attacks, blue team responds\n- **Iterative Hardening**: Multiple rounds of attack-defend-improve\n- **Cross-Pollination**: Red team insights inform blue team designs\n- **Threat Modeling**: Systematic analysis of attack surfaces\n- **Resilience Testing**: Measure recovery from successful attacks\n\n### Strategic Synthesis\n- **Antifragility Design**: Solutions that get stronger from attacks\n- **Adaptive Immunity**: Learning from attacks to prevent similar ones\n- **Offensive Defense**: Using defensive insights for offensive advantages\n- **Meta-Security**: Protecting the security system itself\n- **Dynamic Equilibrium**: Balanced offensive-defensive capabilities\n\n## Battle-Tested Output Format\nProvide a war-gamed solution with:\n- **Attack Surface Analysis**: Comprehensive vulnerability assessment\n- **Defensive Architecture**: Multi-layered protection strategies\n- **Battle-Test Results**: How the solution performed under attack\n- **Adversarial Intelligence**: Insights gained from the red-blue exercise\n\nOnly provide the output in the following JSON format enclosed in <json> tags:\n\n<json>\n{\n\"initial_prompt_evaluation\": \"Red team analysis identifying vulnerabilities, weak assumptions, and attack vectors in the original prompt\",\n\"refined_prompt\": \"Battle-hardened prompt with defensive measures, redundancy, and antifragile design elements\",\n\"explanation_of_refinements\": \"Detailed explanation of adversarial techniques applied, vulnerabilities addressed, and how the prompt now withstands strategic attacks\"\n}\n</json>",
  "examples": [
    "Design a secure authentication system",
    "Create a business strategy for market entry", 
    "Develop a research methodology for controversial topics",
    "Build a crisis response plan"
  ]
}