literature-survey
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a susceptibility to indirect prompt injection because it is designed to autonomously fetch, process, and summarize content from uncontrolled external sources.
- Ingestion points: External data enters the agent context through several tools, including
brave_web_search,google-scholar-scraper,get_paper_content(from arXiv via alphaxiv), and various Semantic Scholar API endpoints. - Boundary markers: The orchestrator does not specify explicit delimiters or instructions to ignore potential commands embedded within retrieved research papers or web pages.
- Capability inventory: The skill has the ability to write findings to context files via
context-checkpointand can perform recursive actions (like citation chaining and further searching) based on the information it ingests. - Sanitization: There are no documented mechanisms for filtering or sanitizing the text retrieved from external APIs or scrapers before it is processed by the agent.
Audit Metadata