eli5
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external content such as research papers, arXiv IDs, and DOIs. This creates a surface for indirect prompt injection, where an attacker could embed hidden instructions within a paper to influence the agent's output.
- Ingestion points: External URLs, paper IDs, and DOI strings processed via the
feynmantool. - Boundary markers: The instructions do not define specific delimiters or guardrails for the external content.
- Capability inventory: The skill utilizes the
feynman alphaCLI tool to fetch or process data. - Sanitization: No explicit content filtering or sanitization steps are documented.
- [COMMAND_EXECUTION]: The skill relies on the execution of a specific command-line tool,
feynman alpha, to retrieve and analyze research data. This tool-specific call is the primary mechanism for the skill's functionality.
Audit Metadata