term-sheet-review
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [Category 1: Prompt Injection] (SAFE): No instruction override, role-play, or system prompt extraction patterns detected.
- [Category 2: Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or network operations found.
- [Category 4: Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill contains no code, package managers, or remote execution patterns.
- [Category 8: Indirect Prompt Injection] (LOW): 1. Ingestion points: User-provided term sheet text. 2. Boundary markers: Absent. 3. Capability inventory: No network, file-write, or shell capabilities. 4. Sanitization: Absent. The vulnerability surface is present but non-exploitable due to the absence of dangerous capabilities.
Audit Metadata