prd
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and provides a template for generating documentation. It does not contain any executable code or instructions that bypass safety filters.
- [DATA_EXFILTRATION]: The skill writes output to a local directory (/tasks/). This is the intended primary purpose of the skill (saving generated documents) and does not involve sending data to external servers or accessing sensitive system files.
- [PROMPT_INJECTION]: No evidence of jailbreaking, system prompt extraction, or instructions to ignore safety guidelines was found. The 'Important' markers in the text are used for valid formatting and process constraints.
- [REMOTE_CODE_EXECUTION]: The skill does not perform any network requests, package installations, or dynamic script executions.
Audit Metadata