skill-creator-ms
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The analyzed content is an instructional guide for creating other skills. It does not contain malicious code, hidden instructions, or exfiltration patterns. It explicitly warns against anti-patterns like hardcoding credentials.
- [EXTERNAL_DOWNLOADS]: The skill references official Microsoft Learn documentation and Azure GitHub repositories as trusted sources for SDK information.
- [INDIRECT_PROMPT_INJECTION]: The skill describes a workflow that ingests user-provided SDK context. Ingestion points: User-supplied SDK names and documentation URLs are requested to guide the creation process. Boundary markers: The instructions encourage the use of clear markdown headers and section delimitation. Capability inventory: As an instructional skill, it does not possess active execution capabilities like shell access or networking itself. Sanitization: Includes a mandatory check for 'Forbidden Patterns' such as API keys and hardcoded credentials in the generated output to ensure safety.
Audit Metadata