The Agent Skills Directory

[SAFE]: The skill serves as a static analysis tool (scripts/skill_security_auditor.py) to help users identify security issues in other extensions. All detected patterns are defensive in nature.
[COMMAND_EXECUTION]: The subprocess.run() usage in scripts/skill_security_auditor.py is safely implemented with list arguments and shell=False to execute standard git clone operations during remote audits.
[PROMPT_INJECTION]: Prompt injection flags in SKILL.md and references/threat-model.md are benign text strings used as descriptive labels or examples of patterns the scanner looks for (e.g., 'Ignore previous instructions'). They do not attempt to manipulate the analyzing agent.
[OBFUSCATION]: The references to base64 and destructive operations in references/threat-model.md are entirely educational documentation detailing known attack types for threat modeling purposes, rather than executable code.

skill-security-auditor