geopolitical-analyst

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external content through user-supplied articles and live web research.
  • Ingestion points: As defined in SKILL.md, the agent processes external articles and performs live web research to answer geopolitical questions.
  • Boundary markers: The instructions provide a structured classification system for articles (e.g., "Object of audit", "Source material") and specific warning labels for claims (e.g., "Unverified", "Misleading"), which help the agent maintain a defensive posture when processing external data.
  • Capability inventory: The agent utilizes web browsing tools to access external websites and verify information.
  • Sanitization: The skill mandates independent evidentiary chains, source triangulation, and disconfirmation strategies in references/source-standards.md, effectively creating a logical filter to identify and flag problematic or malicious content within ingested data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 11:32 AM
Security Audit — agent-trust-hub — geopolitical-analyst