skills/mukul975/anthropic-cybersecurity-skills/detecting-business-email-compromise/Gen Agent Trust Hub
detecting-business-email-compromise
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze external email data, which creates a potential surface for indirect prompt injection if the ingested text contains malicious instructions aimed at the AI agent.
- Ingestion points: scripts/agent.py parses email files and scripts/process.py parses email data from JSON.
- Boundary markers: The skill does not implement specific delimiters or instructions to prevent the agent from following commands embedded in the data.
- Capability inventory: All analysis is restricted to local regex processing; the scripts lack network, file system write, or command execution capabilities.
- Sanitization: Data is processed for pattern matching but not sanitized to prevent downstream injection vulnerabilities.
- [SAFE]: The detected risk is inherent to the skill's primary function of cybersecurity analysis, and no evidence of malicious intent or behavior was found in the code.
Audit Metadata