geolocation-skill

Warn

Audited by Gen Agent Trust Hub on Feb 15, 2026

Risk Level: MEDIUMPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • PROMPT_INJECTION (MEDIUM): The skill is susceptible to indirect prompt injection via untrusted external inputs.
  • Ingestion points: Processes user-provided images in scripts/examine_region.py and web search/fetch results in SKILL.md.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the reasoning workflow.
  • Capability inventory: The skill possesses file-writing capabilities (saving processed images) and network access (WebSearch/WebFetch tools).
  • Sanitization: No logic exists to sanitize or filter text extracted from images or external web pages before processing.
  • EXTERNAL_DOWNLOADS (LOW): The script scripts/examine_region.py requires the Pillow library. While this is a standard image processing package, it is an external dependency and no version constraints are provided in the skill documentation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 15, 2026, 11:22 PM