Thesis-WireMCP_server

ADourou/Thesis-WireMCP_server

3.1

If you are the rightful owner of Thesis-WireMCP_server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The project involves developing an MCP server that uses AI to translate natural language into tshark commands for advanced pcap analysis.

🎓 My Thesis Project - Wireshark MCP server

Status

Thesis: An MCP server (Node.js) where an AI autonomously translates natural language into dynamic tshark commands for pcap analysis.

💡 Project Motivation

This project extends the analyze_pcap tool found in the WireMCP project.

While WireMCP provides an excellent foundation for MCP-based analysis, its default tool is limited to basic 5-tuple fields (e.g., ip.src, ip.dst, tcp.srcport). This is insufficient for the more demanding queries required by academic scenarios (like the Jim Kurose Wireshark Labs), which need access to a wide variety of specific protocol fields (e.g., http.host, dns.cname).

✨ My Solution

The core of my project is an MCP server where the AI client autonomously translates a user's natural language question (e.g., "Find all CNAMEs for www.mit.edu") into the exact tshark display filters (-Y) and fields (-e) needed to get the answer.

This moves beyond fixed queries and allows for truly dynamic and deep packet analysis.

  • Repository: The code is currently in a private repository pending completion and review.

🎓 Academic Info

  • Author: Anastasia Dourou
  • Institution: University of Crete
  • Department: Computer Science Department
  • Supervisor: Xenofontas Dimitropoulos