The Agent Skills Directory

[Prompt Injection] (SAFE): No malicious instructions or bypass attempts were detected in the skill's documentation or templates.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or sensitive file access patterns were found. All networking patterns in templates are directed toward local infrastructure like Redis.
[Indirect Prompt Injection] (SAFE): The skill processes user data as part of its memory management and caching logic.
Ingestion points: query and messages parameters in the cached_search and cached_add functions within templates.
Boundary markers: None are present in the provided configuration templates.
Capability inventory: Local memory operations including search, add, update, and deletion.
Sanitization: User content is passed directly to the memory library without additional sanitization layers.
[Dynamic Execution] (SAFE): The skill generates local Python scripts and configuration files from hardcoded templates. These operations are local, deterministic, and do not involve executing untrusted remote code.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill uses well-known, legitimate packages and does not attempt to download or execute external scripts at runtime.

memory-optimization