PPhying中文
Curated security research, vulnerabilities, advisories and tools for practitioners.

Exploring Large Language Models: Local LLM CTF & Lab

bishopfox.com | blog | #ai-security | #ctf | #jailbreak | #llm-security | #prompt-engineering | #llm-quarantine
📄 Raw .md⬇️ Download .md🔗 Original ↗

Summary

Bishop Fox builds a local LLM capture-the-flag: a controller enforces constraints between user input and a privileged model, showing how crafted questions pass semantic checks to extract its secret.
Published
Collected

Related coverage

blog · netspi.com
Balancing Security and Usability of Large Language Models: An LLM Benchmarking Framework
NetSPI's open LLM security benchmark tests how large language models resist jailbreaks under adversarial conditions while measuring the security-usability trade-off in enterprise deployments.
blog · josephthacker.com
Jailbreaking Humans vs Jailbreaking LLMs
rez0 argues that jailbreaking an LLM mirrors social-engineering humans: both fixate on recent input without full context. He predicts near-infinite context windows could make jailbreaks far harder.
tool · github.com
LLM SecRange: Open-Source Penetration Testing Playground for Large Language Models & Vulnerable Agents
Offers security researchers and penetration testers an out-of-the-box, comprehensive environment to practice prompt injection, OWASP LLM Top 10 risks, and agent tool-abuse exploitation against real models.
blog · bishopfox.com
Ready to Hack an LLM? Our Top CTF Recommendations
Curated LLM security CTFs and playgrounds: Bishop Fox's Local LLM CTF, OWASP FinBot, Steve's Chat Playground, PortSwigger's Web LLM Attacks, and Wild LLaMa cover prompt injection and jailbreaks.
blog · bishopfox.com
Broken Hill: A Productionized Greedy Coordinate Gradient Attack Tool for Use Against Large Language Models
Bishop Fox explains the GCG attack behind adversarial LLM prompts and introduces Broken Hill, which generates jailbreak strings for popular models — runnable locally on GPUs like an RTX 4090.
blog · portswigger.net
Functional PoCs in less than a minute? Julen Garrido Estévez puts Burp AI to the test
Pentester Julen Garrido Estévez systematically benchmarks Burp AI on Web Security Academy labs, comparing prompt styles and measuring time-to-PoC, cost, and hallucination rates.
Back