skills/sorawit-w/agent-skills/storm/Gen Agent Trust Hub

storm

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions focus on retrieval-grounded research and include significant integrity mechanisms. The core logic relies on a structural 'source-to-claim' map designed to prevent hallucinations and identify unverified information.
  • [PROMPT_INJECTION]: The skill processes untrusted external data during its retrieval phase, creating a surface for indirect prompt injection.
  • Ingestion points: Retrieval cycles in Phase 2 documented in SKILL.md and references/loop-mechanics.md.
  • Boundary markers: Implements structural source-to-claim mapping and mandatory [unverified] tagging for any content not directly backed by retrieval results as specified in references/honesty-rubric.md.
  • Capability inventory: Utilizes available platform retrieval tools (e.g., search, fetch) to gather external knowledge.
  • Sanitization: Uses structural attribution and dedicated 'rigor passes' (contradiction mapping and peer review) to evaluate the reliability and bias of external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 07:12 AM
Security Audit — agent-trust-hub — storm