literature-review
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches research papers, metadata, and abstracts from multiple external academic APIs, including PubMed, arXiv, bioRxiv, and Semantic Scholar using the requests library.
- [COMMAND_EXECUTION]: It uses the Bash tool to execute local Python scripts (such as search_databases.py, verify_citations.py, and generate_pdf.py) and system-level tools like pandoc and xelatex for data aggregation and document generation.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to the ingestion of untrusted content from external academic sources into the agent's context for summarization.
- Ingestion points: Retrieval of paper titles and abstracts via external JSON results processed by internal scripts.
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions to isolate external research text from agent directives.
- Capability inventory: Access to the Bash tool for script execution and file system tools (Read, Write, Edit).
- Sanitization: There is no specified logic for sanitizing or validating external literature content before it is processed by the model.
Audit Metadata