tax-fraud-detection
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE] (SAFE): The skill consists entirely of Markdown text, instructions, and tables. There are no scripts, binaries, or command-line instructions provided within the file.
- [EXTERNAL_DOWNLOADS] (SAFE): No external URLs, repositories, or package managers (such as npm or pip) are referenced or used.
- [DATA_EXFILTRATION] (SAFE): The skill lacks any networking capabilities, file system access, or hardcoded credentials. It does not contain commands like curl, wget, or fetch.
- [PROMPT_INJECTION] (SAFE): No instructions were found that attempt to override the AI agent's core safety guidelines, bypass content filters, or extract the system prompt. The focus is exclusively on legitimate tax audit procedures.
Audit Metadata