analyzing-research-papers
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection (LOW): The skill methodology involves reading and analyzing untrusted research papers from URLs and local file paths. This establishes a surface where malicious instructions could be embedded within the paper's content to influence the agent's behavior. Ingestion points: Identified as external URLs (ArXiv, journal websites), DOI metadata, and local file paths. Boundary markers: The instructions do not define specific delimiters or 'ignore embedded instructions' warnings to isolate external data from the agent's core logic. Capability inventory: The skill allows for comprehensive technical analysis, data extraction, and structured reporting based on external text. Sanitization: There are no provisions for input validation, escaping, or filtering of the retrieved research content.
Audit Metadata