MCP Security Testing

In this module of the Advanced AI Security Part 1 learning path, you will practice MCP security testing against the MathGPT MCP server as an API. The workspace rule is documentation. read_workspace resolves paths without a jail while guardrails are off. Upon completion of the module, students will be able to: - Call tools/list and tools/call with hostile paths. - Show that list_workspace is not a security boundary. - Write the finding as path traversal, not as an LLM jailbreak. For inquiries, please write us at https://www.darkrelay.com/cybersecurity-course-inquiry