opportunity-tool-selection
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of high-level logical instructions for tool routing and contains no executable code or scripts.
- [SAFE]: The instructions include a dedicated 'Boundaries' section that explicitly prohibits the agent from attempting to bypass security controls such as MFA, CAPTCHAs, and payment walls.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a framework for ingesting external data (Ingestion) from web searches, browser automation, and local files. It identifies capabilities like MCP/API access to systems of record. While specific sanitization is not detailed, the skill provides specific boundary markers (Boundaries) to mitigate the risk of obeying malicious instructions embedded in external content.
Audit Metadata