The Agent Skills Directory

[Prompt Injection] (SAFE): No malicious instructions, jailbreak attempts, or safety bypass markers detected. While the metadata description uses the keyword 'CRITICAL', it is contextually used as a trigger for documented Rust compiler errors (E0596, E0499, E0502) rather than an attempt to override the agent's system instructions.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or network operations (curl, wget, fetch) are present.
[Remote Code Execution] (SAFE): The skill does not download or execute remote scripts. No 'npm install' or 'pip install' commands are present.
[Command Execution] (SAFE): No shell commands, subprocess calls, or administrative privilege escalation attempts were found.
[Obfuscation] (SAFE): No Base64, zero-width characters, or homoglyph-based encoding techniques are used to hide content.
[Indirect Prompt Injection] (SAFE): The skill does not ingest untrusted data or provide surfaces for external interpolation. It lacks capability for file-writing or network interaction, eliminating the possibility of tool output poisoning.

m03-mutability