swe-test-engineer
Warn
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a template in references/export.md to dynamically generate a Python script and then executes it using python3 to create Excel spreadsheets.\n- [EXTERNAL_DOWNLOADS]: During the export process, the skill attempts to install the openpyxl package from the Python Package Index (PyPI) using pip if it is not already available on the system. Documented as a well-known service download.\n- [PROMPT_INJECTION]: The skill processes untrusted user-provided requirements and incorporates them into a generated Python script that is subsequently executed, creating an indirect injection surface.\n
- Ingestion points: User-provided test requirements and field constraints processed in SKILL.md and related reference files.\n
- Boundary markers: None identified. The instructions do not specify delimiters or warnings to ignore embedded instructions in the requirement text.\n
- Capability inventory: The skill has the capability to generate and execute arbitrary Python scripts and create files/directories as documented in references/export.md.\n
- Sanitization: The skill instructions recommend slugifying the requirement name for the filename, but do not specify comprehensive sanitization for the data being embedded into the generated script.
Audit Metadata