geo-leaderboard

Pass

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external AI providers to extract brand names and perform sentiment analysis.
  • Ingestion points: AI model responses generated during the leaderboard execution phase (Step 5).
  • Boundary markers: A manual review step is present for generated queries in Step 4, but no specific boundary markers or validation steps are defined for the responses received from AI providers.
  • Capability inventory: The skill utilizes the Bash tool to run command-line analysis and the Read/Write tools to handle query and result files.
  • Sanitization: No explicit sanitization or filtering of external AI content is described before the extracted data is processed and presented to the user.
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute the voyage_geo command-line interface. This tool is used to manage providers, generate queries, and run the leaderboard analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 27, 2026, 02:38 PM