building-sales-team

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): No patterns of direct prompt injection or safety bypass were detected. An indirect prompt injection surface exists where the agent ingests user-provided business metrics, but this is assessed as safe because the skill has no functional capabilities (scripts, tools, or network access) that could be exploited. Evidence: 1. Ingestion points: User-provided responses to sales growth questions (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: None. 4. Sanitization: N/A.
  • [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file path references, or network transmission commands were found.
  • [Remote Code Execution] (SAFE): The skill contains no external dependencies or instructions for downloading/executing remote scripts.
  • [Obfuscation] (SAFE): The markdown content is clear and readable, with no evidence of Base64 encoding, zero-width characters, or homoglyphs.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 04:44 PM