deep-reading-analyst

Pass

Audited by Gen Agent Trust Hub on Mar 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional and document-based. It guides the AI agent to apply various cognitive frameworks to user-provided content.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external content (articles and research papers). This creates a surface for indirect prompt injection where instructions embedded in the processed text might attempt to influence the agent's behavior.
  • Ingestion points: User-provided URLs and long-form content processed in SKILL.md.
  • Boundary markers: The skill uses structural templates and specific framework questions which serve as natural constraints, though no explicit technical delimiters are defined for external content.
  • Capability inventory: The skill utilizes web_search for research (Level 4).
  • Sanitization: No specific sanitization or filtering of external content is implemented; the agent relies on its internal safety guidelines and the structured analysis process defined in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 17, 2026, 08:29 PM