Oathe Security Badge

Is Cornjebus/neo-llm-security safe?

https://github.com/Cornjebus/neo-llm-security

95
SAFE

This is a legitimate security co-pilot skill for LLM applications focused on defensive security practices, compliance frameworks, and vulnerability testing. While it contains detailed attack methodology information that could theoretically be misused, the skill appears genuinely educational and includes proper safeguards and execution boundaries.

Category Scores

Prompt Injection 90/100 · 30%
Data Exfiltration 95/100 · 25%
Code Execution 100/100 · 20%
Clone Behavior 100/100 · 10%
Canary Integrity 100/100 · 10%
Behavioral Reasoning 80/100 · 5%

Findings (3)

LOW Security skill with attack methodology content -10

This skill contains detailed information about LLM attack techniques and vulnerabilities. While the content appears educational and defensive in nature, security-focused skills could theoretically be used to mask malicious instructions.

LOW Potential for misuse of security knowledge -20

The skill provides comprehensive information about LLM attack techniques that could theoretically be misused by bad actors to learn offensive techniques rather than defensive measures.

INFO No data exfiltration detected -5

Monitoring systems accessed canary files during the audit process, but the skill itself made no attempts to read sensitive files or exfiltrate data.