travel-skill

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses a set of local Python scripts (scripts/) to perform linting and validation of film production data. These scripts use standard libraries (pathlib, json, re, argparse) and contain no network operations, file-system writes, or dangerous execution patterns.
  • [SAFE]: No prompt injection or jailbreak attempts were found. The instructional language is strictly professional, focusing on cinematography, lighting, and spatial logic.
  • [SAFE]: No data exfiltration or credential harvesting patterns were identified. The skill does not access sensitive directories (e.g., .ssh, .aws) or use networking tools like curl or requests.
  • [SAFE]: No obfuscation techniques, such as Base64-encoded commands or hidden Unicode characters, are present in the scripts or reference documents.
  • [SAFE]: The skill manages indirect prompt injection risks by utilizing structured templates (YAML/Markdown) and requiring spatial audits of source frames. While it processes user-provided script data, the surface area is limited to local validation without downstream execution of untrusted input.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 06:08 PM
Security Audit — agent-trust-hub — travel-skill