The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data, such as game tool descriptions and file paths, which presents a surface for indirect prompt injection.
Ingestion points: The analyze_cheat_features(readme_text) function ingests arbitrary text descriptions, and static_analysis_safe(file_path) accepts user-specified file paths.
Capability inventory: The skill possesses the ability to execute system commands via subprocess.run and monitor system processes via psutil.
Boundary markers: There are no explicit delimiters or instructions for the agent to ignore embedded commands within the analyzed text.
Sanitization: The skill performs basic normalization (lowercasing) for keyword detection but lacks formal sanitization for file paths passed to system utilities.
[COMMAND_EXECUTION]: The skill implements a static analysis function that invokes system-level commands.
Evidence: The static_analysis_safe function uses subprocess.run(['strings', file_path], ...) to extract strings from binary files. While the use of a list for arguments mitigates typical shell injection, it remains a system command execution based on external input.

game-cheat-detection-and-security-analysis