The Agent Skills Directory

[SAFE]: The skill serves as a static evaluation framework. It provides detailed metrics and procedures for human or agent-driven auditing of SKILL.md files. No malicious code, exfiltration patterns, or obfuscation were found in the provided files.
[INDIRECT_PROMPT_INJECTION]: The skill is intended to process external, untrusted skill files. While this creates a surface for indirect prompt injection, the skill's protocol emphasizes critical analysis and dimension-based scoring rather than execution of instructions within the target files. This analytical posture naturally mitigates the risk of following adversarial instructions embedded in audited content.
[NO_CODE]: The skill contains only documentation and instructional prompts. It does not utilize any scripts, executables, or network-capable tools, which minimizes the attack surface and prevents common automated exploit vectors.

skill-judge