blur-image
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the local
magickcommand-line utility for metadata extraction, calibration grid drawing, and Gaussian blurring. These operations are restricted to image processing. - [PROMPT_INJECTION]: The skill processes untrusted data via multimodal image analysis. Ingestion point: Phase 2, Step 1 in SKILL.md. Boundary markers: None. Capability inventory: Local command execution (
magick) in Phase 1 and 4. Sanitization: The skill employs coordinate-based redaction and a verification phase to ensure the agent correctly identifies and obscures sensitive content.
Audit Metadata