Claudeisms Operational Guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [PROMPT_INJECTION] (SAFE): No instructions attempt to bypass safety filters, extract system prompts, or override core agent constraints. The guidelines focus on task formatting and operational logic.- [DATA_EXFILTRATION] (SAFE): No network operations, hardcoded credentials, or sensitive file path access were detected.- [COMMAND_EXECUTION] (SAFE): While the skill mentions bash commands and database queries in its guidelines (e.g., 'Minimal timeout for bash commands'), it does not provide or execute specific malicious commands. It encourages safety checks like confirmation before deletion.- [EXTERNAL_DOWNLOADS] (SAFE): The skill does not define any external package dependencies or remote code sources.- [OBFUSCATION] (SAFE): No encoded content, zero-width characters, or homoglyphs were found.
Audit Metadata