analyzing-credit-agreement-terms
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted documents such as credit agreements and term sheets. This constitutes an indirect prompt injection attack surface where malicious instructions could be embedded in the analyzed text to influence the agent's behavior. However, this is inherent to the skill's primary purpose of document analysis.
- [NO_CODE]: The skill consists entirely of instructional markdown and YAML metadata. It does not include any scripts, executables, or configurations that perform active operations on the host system.
Audit Metadata