han-river-water-level

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses curl to perform HTTP GET requests to retrieve hydrological data and mentions jq for JSON processing.
  • [EXTERNAL_DOWNLOADS]: Hydrological data is downloaded from the external proxy server k-skill-proxy.nomadamas.org.
  • [DATA_EXFILTRATION]: The skill transmits station names and codes to k-skill-proxy.nomadamas.org. This network communication is required for the skill's primary functionality and does not involve sensitive system files.
  • [PROMPT_INJECTION]: The skill implements an indirect prompt injection surface by processing content from an external API response for summarization without sanitization.
  • Ingestion points: API responses from https://k-skill-proxy.nomadamas.org/v1/han-river/water-level.
  • Boundary markers: None present in the instructions to separate API data from agent instructions.
  • Capability inventory: Uses curl for network requests.
  • Sanitization: No validation or sanitization of external API data is performed before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 01:00 AM
Security Audit — agent-trust-hub — han-river-water-level