product-skills
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: All core logic is implemented in local Python scripts that rely exclusively on the standard library. No external network connections, sensitive file system access, or suspicious data operations were found.
- [SAFE]: The skill architecture is transparent and follows its stated purpose of routing and validating product discovery work without any signs of obfuscation, persistence, or privilege escalation.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user goals for classification and routing. This ingestion surface is managed by a keyword-based deterministic script (
product_goal_router.py), and the resulting actions are restricted to pre-defined product management sub-skills, minimizing potential exploitation risks.
Audit Metadata