asta-corpus-search
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes scientific paper metadata and full-text passages, creating a potential surface for indirect prompt injection attacks where instructions could be embedded in the retrieved content.
- Ingestion points: Data processed through the tools search_papers, get_papers, and get_citations described in SKILL.md.
- Boundary markers: The skill does not define specific boundaries or instructions to differentiate between the agent's instructions and the external data being processed.
- Capability inventory: Capabilities are limited to data retrieval via network requests. No high-risk tools such as shell execution or file system writes are present.
- Sanitization: No content filtering or sanitization mechanisms are specified for the incoming data.
- [EXTERNAL_DOWNLOADS]: The skill interacts with established scientific literature platforms.
- Details: It facilitates communication with APIs from Allen AI (asta-tools.allen.ai), PubMed (ncbi.nlm.nih.gov), OpenAlex (api.openalex.org), and Semantic Scholar (api.semanticscholar.org).
Audit Metadata