clade-policy-guardrails
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides instructional patterns for building safer applications, specifically focusing on prompt injection defense and system prompt hardening.
- [SAFE]: Includes TypeScript code examples for input and output validation that scan for common injection techniques such as 'DAN' or instructions to ignore previous rules.
- [SAFE]: References official documentation from Anthropic for safety best practices.
- [SAFE]: No evidence of data exfiltration, malicious command execution, or obfuscation was found in the skill content or its metadata.
Audit Metadata