PatchIT aims to enhance the performance of its AI agents in handling cybersecurity queries, particularly around CVE data. The project involves prompt engineering, testing, and refining the interaction logic of these AI agents to improve their accuracy and efficiency. Students will design effective instructions, test the agents' responses, and refine the logic to ensure better handling of CVE-related queries. The project offers a hands-on opportunity to apply classroom knowledge in AI and cybersecurity, focusing on real-world scenarios and data. The ultimate goal is to streamline the agents' responses and automate repetitive queries, thereby improving overall system performance.

- Understand CVE analysis and their current capabilities.

- Design and test various prompts for CVE queries.

- Refine the agents' response logic for better accuracy.

- Automate common analysis patterns for repetitive tasks.

- Document the process and provide recommendations for future improvements.

The project will result in a set of refined instructions and improved interaction logic for our AI agents, specifically for CVE analysis. Students will also create templates for automating repetitive queries and provide comprehensive documentation of their methodologies and findings.

- Refined instructions for CVE-related queries.

- Improved interaction logic for AI agents.

- Templates for automating repetitive queries.

- Detailed documentation and a final report with recommendations.