Plugin & Extension Security for AI: Complete Guide | QuizBy Eyal Doron / December 6, 2025 / 1 minute of reading Plugin & Extension Security for AI: Complete Guide | Quiz 1 / 7 1. According to the article – why is the misconception that plugins just read data not harmful dangerous? 1. Data exfiltration is impossible through plugin architecture 2. Many plugins have write and execute capabilities and even read-only plugins can exfiltrate data 3. All plugins are strictly read-only by default 4. Reading data requires special permissions that are never granted Correct! Why: Many plugins have write and execute capabilities and even read-only plugins can exfiltrate sensitive data – read access to the wrong data is still a breach. Context: The assumption that reading is harmless ignores both hidden capabilities and data sensitivity. Remember: Read access to secrets is still a breach. 2 / 7 2. What security testing framework does the article recommend before approving plugins? 1. Static analysis – dynamic analysis – integration testing – permission testing 2. User acceptance testing in production 3. Antivirus scanning of plugin files 4. Only manual code review by developers Correct! Why: The article recommends static analysis for code vulnerabilities plus dynamic analysis for runtime behavior plus integration testing for secure channels plus permission testing for least privilege. Context: This comprehensive testing catches different types of security issues before plugins enter production. Remember: Static – Dynamic – Integration – Permission testing. 3 / 7 3. What guidance does the article give for Anthropic Claude MCP security? 1. Use only public MCP endpoints for convenience 2. Disable MCP entirely as it cannot be secured 3. Host MCP servers internally – authenticate via mTLS or API keys – monitor unusual tool sequences 4. Allow all MCP tools without restriction Correct! Why: The article recommends hosting MCP servers internally and authenticating calls via mTLS or API keys and monitoring for unusual tool sequences. Context: Local MCP servers have file system access to the host machine making careful implementation review essential. Remember: Internal hosting – strong auth – watch for anomalies. 4 / 7 4. What are the four defensive layers for plugin security described in the article? 1. Firewall – antivirus – encryption – backup 2. Prevention – detection – response – recovery 3. Vetting and approval – least privilege – sandboxing and isolation – monitoring and auditing 4. Authentication – authorization – accounting – auditing Correct! Why: The article describes vetting and approval followed by least privilege followed by sandboxing and isolation followed by monitoring and auditing as the four layers. Context: These layers work together to prevent bad plugins from entering and contain damage when things go wrong. Remember: Vet – Limit – Contain – Watch. 5 / 7 5. What is OWASP LLM #7 and why does it exist? 1. Insecure Plugin Design – exists because plugin security is systematically weak across the industry 2. An encryption protocol for plugin communications 3. A certification standard for secure plugin development 4. A specific type of malware that targets AI systems Correct! Why: OWASP LLM #7 addresses Insecure Plugin Design – it exists because plugin security is systematically weak across the industry. Context: Plugins often receive excessive permissions and lack input validation because they are built quickly with security as secondary. Remember: Assume plugins are insecure until proven otherwise. 6 / 7 6. What are the three main plugin threat vectors identified in the article? 1. Encryption – hashing – tokenization 2. Malicious plugins – compromised plugins – exploited plugins via prompt injection 3. Authentication – authorization – accounting 4. Network attacks – application attacks – physical attacks Correct! Why: The article identifies plugins that can be the attacker (malicious) or the victim (compromised) or the weapon (exploited via prompt injection). Context: Defense must address all three vectors because each requires different security controls. Remember: Malicious – Compromised – Exploited. 7 / 7 7. According to the article – what analogy best describes the security risk of AI plugins? 1. Giving your AI keys to different rooms in your building – each key could be misused 2. Creating backup copies of important documents 3. Building a firewall around your network perimeter 4. Installing antivirus software on multiple computers Correct! Why: The article describes plugins as giving your AI assistant keys to different rooms in your building – each key could be misused. Context: This analogy illustrates that every plugin grants access to systems and data that could be exploited if the AI is manipulated or the plugin is malicious. Remember: Every plugin is a key that could end up in the wrong hands. Your score isThe average score is 0% Restart quiz Download PDF Please leave this field empty๐ The AI Security Manager's Newsletter Weekly insights on AI risk management, EU AI Act compliance, and practical security strategies. We donโt spam! Read our privacy policy for more info. Thank you! Please check your inbox to confirm your subscription.
