Skip to main content

adversarial-examples

Generates adversarial inputs and edge cases to stress-test LLM robustness, enhancing model reliability and performance.

Install this skill

or
99/100

Security score

The adversarial-examples skill was audited on Mar 3, 2026 and we found 1 security issue across 1 threat category. Review the findings below before installing.

Categories Tested

Security Issues

low line 163

Unicode escape sequences

SourceSKILL.md
163ZERO_WIDTH = ['\u200b', '\u200c', '\u200d', '\ufeff']
Scanned on Mar 3, 2026
View Security Dashboard
Installation guide →