bug-bounty541
xss375
exploit251
google199
rce181
facebook153
microsoft126
web3124
writeup121
open-source91
csrf88
cve84
account-takeover79
apple76
browser71
malware65
sqli64
ai-agents63
ssrf54
tool46
dos45
oauth44
privacy44
pentest39
supply-chain39
idor39
privilege-escalation38
ctf38
llm37
lfi36
opinion35
auth-bypass34
cors34
react33
phishing33
reverse-engineering33
automation33
machine-learning32
aws32
code-generation31
infrastructure31
clickjacking31
race-condition30
cloudflare30
cloud30
node29
access-control27
subdomain-takeover26
rust24
performance-optimization24
0
1/10
humor
A GitHub gist documenting humorous and nonsensical LLM responses, including examples of AI assistants failing to find bugs, contradicting themselves, and generating invalid commands — primarily showcasing failures of LLMs like Claude and ChatGPT when given edge-case or conflicting prompts.
OpenAI
Claude
Anthropic
Matt Pocock