pricing
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: Analysis of the skill files found no evidence of malicious patterns, data exfiltration, or unauthorized command execution. The skill is entirely composed of static documentation and evaluation benchmarks.
- [PROMPT_INJECTION]: The skill processes untrusted business context from external files like .agents/product-marketing.md. While this is a vulnerability surface for indirect prompt injection, it is mitigated by the fact that the skill has no dangerous capabilities or tools to exploit. Ingestion points: .agents/product-marketing.md and .claude/product-marketing.md as referenced in SKILL.md. Boundary markers: None. Capability inventory: No high-risk tools (network, shell, or file-write) are requested or used. Sanitization: Not implemented.
Audit Metadata