saas-msa-review
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is strictly instructional and contains no executable code or network operations. Analysis found no evidence of obfuscation, credential theft, or unauthorized persistence.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted legal text for analysis. While this represents an injection surface, the risk is assessed as safe because the skill has no executable capabilities (such as network access, shell command execution, or file writing) that could be manipulated by an attacker. 1. Ingestion points: Contract text provided by the user. 2. Boundary markers: None identified. 3. Capability inventory: None. 4. Sanitization: None.
Audit Metadata