mat-dft-vasp
Pass
Audited by Gen Agent Trust Hub on Jun 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides Python scripts to automate the creation of VASP input files (INCAR, POTCAR, KPOINTS, POSCAR) and the parsing of output files (vasprun.xml, OUTCAR). These tools follow standard scientific workflows and utilize well-known libraries such as ASE and Pymatgen.
- [COMMAND_EXECUTION]: The provided scripts perform file system operations to read structure files and write simulation inputs. These operations are restricted to paths provided by the user and do not involve the execution of arbitrary shell commands or privileged operations.
- [DATA_EXFILTRATION]: No network operations or attempts to access sensitive configuration files (e.g., SSH keys, AWS credentials) were found. The skill processes local scientific data files.
- [SAFE]: The skill processes untrusted external data (CIF/XYZ structures and XML outputs). The risk of indirect prompt injection is mitigated by the use of established scientific parsing libraries which validate the data schema.
- Ingestion points: Structure files (CIF, XYZ, POSCAR) and VASP XML/OUTCAR files.
- Boundary markers: Not explicitly required as input is handled by library-level parsers.
- Capability inventory: Local file system writes to create simulation input directories.
- Sanitization: Data is validated and converted into structured objects by the ASE and Pymatgen libraries before further processing.
Audit Metadata