contractor-agreement-review

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOWPROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection (LOW): The skill analyzes untrusted documents provided by users, which creates a surface for indirect prompt injection where a document could contain instructions intended to mislead the agent's analysis.
  • Ingestion points: User-provided contractor agreements analyzed in CHECK and ADVISE modes (SKILL.md).
  • Boundary markers: Absent. The instructions do not specify delimiters to isolate untrusted text from the logic.
  • Capability inventory: None. The skill only performs text-based analysis and returns formatted markdown reports. No subprocess calls, network requests, or file writes are present in any script or instruction.
  • Sanitization: None detected. The skill performs direct analysis of input content against local reference files.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 01:49 AM