The Agent Skills Directory

[Prompt Injection] (SAFE): No attempts to override agent behavior, bypass safety guidelines, or extract system prompts were detected.
[Data Exposure & Exfiltration] (SAFE): There are no hardcoded credentials, sensitive file path references, or network-based data transmission patterns.
[Obfuscation] (SAFE): The file contains plain text documentation with no evidence of Base64 encoding, zero-width characters, or homoglyph-based evasion.
[Remote Code Execution] (SAFE): No remote scripts are downloaded or executed. The skill lacks any runtime code or package management files.
[Persistence & Privilege Escalation] (SAFE): No commands for modifying system startup files, shell profiles, or acquiring elevated permissions (e.g., sudo) are present.

keyboard-navigation-tester