scholar-search
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses the
uvxtool runner to dynamically fetch and execute Python packages (s2cli,openalexcli,arxivy,dblpcli) from the Python Package Index (PyPI). This is a common practice for running portable CLI tools. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external scholarly APIs, creating a surface for indirect prompt injection through untrusted content such as paper abstracts or titles.
- Ingestion points: API responses from Semantic Scholar, OpenAlex, arXiv, and DBLP (SKILL.md).
- Boundary markers: None identified; the agent processes the raw output from the CLI tools.
- Capability inventory: The agent can execute shell commands via
uvxand write to the local file system (e.g., using> refs.bib). - Sanitization: No specific content sanitization or validation of the scholarly metadata is performed before it enters the agent's context.
Audit Metadata