The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted documents provided by users or accessed via tools like read. This creates a surface where instructions embedded in source texts could attempt to influence the agent's behavior. However, the skill explicitly mandates that 'Translation is translation, not explanation' and instructs the agent to 'Not add explanation' and 'Maintain original meaning', which serves as a functional boundary against such injections.
[DATA_EXPOSURE]: No sensitive file paths, hardcoded credentials, or unauthorized network exfiltration patterns were detected. The skill uses standard tools for its intended purpose of document translation.
[COMMAND_EXECUTION]: The skill does not contain any arbitrary command execution, privilege escalation (sudo), or persistence mechanisms.

multi-language