triage-nda

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFE
Full Analysis
  • [Indirect Prompt Injection Surface]: This skill analyzes documents from external sources, which could potentially contain text intended to influence the agent's output or classification logic. Ingestion points: NDA documents are accepted via file upload, URL, or pasted text as defined in SKILL.md. Boundary markers: There are no specific delimiters defined to separate the untrusted document content from the internal analysis instructions. Capability inventory: The skill's functions are limited to text analysis and reporting, with no access to sensitive system commands or network-based exfiltration. Sanitization: No automated sanitization or filtering of the input text is implemented within the provided instructions.
  • [Sensitive Data Processing]: The skill's primary function involves handling confidential legal agreements. Users should ensure that the agent environment adheres to their organization's data privacy and security policies regarding document handling.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 11:58 PM