The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized behaviors were detected. All instructions are consistent with the skill's stated purpose of facilitating audio voice conversion.
[CREDENTIALS_UNSAFE]: The skill includes a proactive security instruction forbidding the collection of passwords, API keys, or access tokens in the chat, directing the agent to use secure OAuth flows instead.
[INDIRECT_PROMPT_INJECTION]: The skill identifies and mitigates potential injection risks from untrusted data (Ingestion: voice names and filenames from tools; Boundary markers: not present; Capability inventory: audio conversion and credit management tools; Sanitization: procedural instruction to treat identifiers strictly as data).

hanolab-mcp