blueprints-maintenance
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No instructions were found that attempt to override system prompts or bypass safety filters.
- [Data Exposure & Exfiltration] (SAFE): The skill performs local file operations on documentation paths and does not attempt to access sensitive system files or communicate with external servers.
- [Indirect Prompt Injection] (SAFE): The skill identifies a surface for indirect prompt injection by reading blueprint files, but since the purpose is strictly documentation maintenance and no dangerous capability chain is present, the risk is classified as safe. Ingestion points: Read("blueprints/..."). Boundary markers: None. Capability inventory: Read, Write, Edit, Grep, Glob. Sanitization: None.
Audit Metadata