deep-research
Pass
Audited by Gen Agent Trust Hub on Jul 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows secure design patterns for AI agents. It explicitly labels external content as data rather than instructions to prevent indirect prompt injection.
- [EXTERNAL_DOWNLOADS]: The skill communicates with established academic infrastructure including Crossref (api.crossref.org), OpenAlex (api.openalex.org), Semantic Scholar (api.semanticscholar.org), and arXiv (export.arxiv.org). These are well-known services used for their intended purpose of bibliographic verification.
- [COMMAND_EXECUTION]: The timeline_extraction_agent utilizes the standard 'pdftotext' command-line utility to extract metadata from local documents. This is a common and legitimate practice for academic research automation.
- [PROMPT_INJECTION]: The skill includes robust internal controls such as the 'Concession Threshold Protocol' in the devils_advocate_agent and 'Intent Detection Layer' in the socratic_mentor_agent to maintain agent integrity and prevent behavioral manipulation.
- [DATA_EXFILTRATION]: While the skill makes external network requests, they are limited to specific academic APIs. No patterns of sensitive data exfiltration (e.g., environment variables or credentials) were observed.
Audit Metadata