skill-catalog-generator
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified during the analysis of the provided files. The skill operates exclusively on repository metadata and documentation files to generate reports.- [PROMPT_INJECTION]: The instructions focus on data normalization and quality scoring without any attempts to override model constraints or safety guidelines.- [DATA_EXFILTRATION]: There are no network operations or commands that would facilitate data transfer to external domains. The skill includes a rule (catalog-gen-5) specifically directing the agent to keep credentials and private data out of public catalogs.- [REMOTE_CODE_EXECUTION]: No remote scripts, package installations, or dynamic code execution patterns were detected.
Audit Metadata