citation-audit

Pass

Audited by Gen Agent Trust Hub on Feb 18, 2026

Risk Level: SAFE
Full Analysis
  • Indirect Prompt Injection (SAFE): The skill identifies a surface for indirect prompt injection due to its core function of processing untrusted external documents (academic papers).
  • Ingestion points: External papers provided by the user or retrieved via the WebFetch tool.
  • Boundary markers: The skill instructions lack explicit XML or unique delimiters for untrusted content, though the highly structured multi-step process and defined output tables act as implicit constraints.
  • Capability inventory: Includes WebSearch, WebFetch, Read, and Write (to output the audit report).
  • Sanitization: No explicit sanitization or instruction to ignore embedded commands within the analyzed papers is present.
  • Data Exposure & Exfiltration (SAFE): The skill uses network-enabled tools (WebSearch, WebFetch) solely for their primary stated purpose of citation verification. No patterns of accessing sensitive local files or hardcoded credentials were found.
  • Unverifiable Dependencies (SAFE): The skill does not define or install external software packages and relies on standard agent tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 18, 2026, 10:39 AM