The Agent Skills Directory

[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or network exfiltration paths were detected. The skill only performs read operations for analysis.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill contains no executable scripts, shell commands, or external package dependencies.
[Indirect Prompt Injection] (SAFE): Although the skill processes external file content (an ingestion surface), it explicitly prohibits code implementation and lacks the capability to write files or access the network, neutralizing potential injection vectors.

decompose-task