The Agent Skills Directory

[SAFE]: The skill instructions and referenced documents contain no malicious code, hidden commands, or security risks.\n- [PROMPT_INJECTION]: The persona instructions are clearly scoped to retail and business advice, with no evidence of jailbreak attempts or safety bypasses.\n- [DATA_EXFILTRATION]: The skill does not perform any network operations or access sensitive local files like credentials or environment variables.\n- [REMOTE_CODE_EXECUTION]: There are no patterns involving the download or execution of external scripts or binaries.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided business questions but lacks exploitable capabilities (such as writing to the filesystem or making network calls) that would enable an indirect prompt injection attack to cause harm.

suzukitoshifumi-perspective