skill-integration
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is defined as 'type: knowledge' and contains no executable scripts or commands. Analysis of the markdown body confirms it is purely instructional content for developers and agents and does not provide any mechanism for malicious execution.
- [SAFE]: No prompt injection patterns were detected. The use of 'FORBIDDEN' and 'REQUIRED' sections serves as a structural guide for skill development and schema enforcement rather than an attempt to override AI safety constraints or agent behavior.
- [SAFE]: The skill does not access sensitive files, hardcode credentials, or perform network operations. It specifies the 'Read' tool in its metadata, which is consistent with its purpose of referencing documentation and example files within the local environment.
- [SAFE]: No obfuscation techniques (such as Base64, zero-width characters, or homoglyphs) or persistence mechanisms were found within the skill's content or metadata.
Audit Metadata