wind-power-business
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted power curve data provided by the user. \n
- Ingestion points: The scripts
compute_cp.pyandfill_cp_workbook.pyread data from user-supplied files or standard input.\n - Boundary markers: There are no formal markers separating user data from the processing logic, although instructions recommend manual parameter verification.\n
- Capability inventory: The skill has file system write access to generate project results and reports.\n
- Sanitization: The processing scripts use specific regular expressions (
NUMBER_RE) to extract only numeric values, which prevents non-numeric malicious instructions from being processed as valid input data.
Audit Metadata