TryHackMe — BankGPT Write-Up
A practical walkthrough of the BankGPT room with focus on prompt injection techniques, LLM security, and context vulnerabilities in banking chatbots.
#security
#ctf
#llm
Timeline
All posts in chronological order.
TryHackMe — EvilGPT V2 Write-Up
A practical guide to prompt injection vulnerabilities in LLMs, demonstrating how system instructions can be revealed through context reinterpretation and structured output formats.
#security
#ctf
#llm