aminer-academic-search
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external content (such as paper abstracts, researcher biographies, and experiment details) retrieved from the AMiner API and incorporates it into the agent's context. This presents a surface for indirect prompt injection if the scholarly data contains malicious instructions designed to influence the agent's behavior.
- Ingestion points: External data enters the context through API responses in
scripts/aminer_client.pyandreferences/api-catalog.md(e.g.,paper_detail,person_detail,experiment_search). - Boundary markers: The instructions do not specify explicit delimiters or guardrails to treat retrieved academic data as untrusted text.
- Capability inventory: The skill has the capability to perform network requests to the vendor's API and execute simple shell commands for environment validation.
- Sanitization: There is no evidence of specific sanitization or filtering of the retrieved natural language data before it is presented to the agent.
Audit Metadata