agently-output-control
Pass
Audited by Gen Agent Trust Hub on Mar 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves its stated purpose of managing structured output formats using the 'agently' library. No malicious intent or suspicious behaviors were identified.
- [PROMPT_INJECTION]: The instructions in
SKILL.mdfocus on guiding the AI toward structured output (JSON/field ordering) and do not contain bypasses, role-play overrides, or instructions to ignore safety protocols. - [DATA_EXFILTRATION]: No evidence of sensitive file access, credential theft, or unauthorized network communication was found. The Python examples focus on local execution and standard framework usage.
- [REMOTE_CODE_EXECUTION]: The skill does not perform any remote code downloads or execution of untrusted scripts. It relies on the 'agently' library, which is consistent with the skill's context.
- [COMMAND_EXECUTION]: The
scripts/check.pyfile performs a benign operation to determine the parent directory name and does not execute arbitrary or dangerous system commands.
Audit Metadata