ase
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill utilizes the
aselibrary, a reputable scientific package from an official registry. All referenced documentation points to well-known academic sources. - [INDIRECT_PROMPT_INJECTION]: The skill enables the ingestion of untrusted external data via file-reading functionality.
- Ingestion points: Usage of
ase.io.read()for various file formats inSKILL.md. - Boundary markers: No specific delimiters or safety instructions are defined for the content of the data files.
- Capability inventory: Capabilities include writing files, configuring tool paths via environment variables, and executing external simulation binaries.
- Sanitization: The
aselibrary parses these files into specific numerical structures, providing structural validation that limits the potential for instruction-based injection.
Audit Metadata