hauling-image-estimator

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were identified. The skill is designed for visual estimation and does not interact with the host system or external networks.
  • [SAFE]: The skill processes user-provided images (ingestion point: SKILL.md), which is a potential surface for indirect prompt injection. However, the risk is negligible as the skill lacks any tools or permissions (such as network access, file-writing, or subprocess calls) that could be exploited. Boundary markers and sanitization are absent, but the lack of capability inventory makes this safe.
  • [SAFE]: No obfuscated content, hardcoded credentials, or suspicious persistence mechanisms were found in the skill's instructions or reference tables.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 05:16 PM
Security Audit — agent-trust-hub — hauling-image-estimator