data-court-auction

Pass

Audited by Gen Agent Trust Hub on Aug 19, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill relies on the installation of external Node.js dependencies to function, specifically the court-auction-notice-search package and optional browser automation tools like playwright-core or rebrowser-playwright. These are purpose-aligned for scraping the target website.
  • [COMMAND_EXECUTION]: The documentation provides instructions for executing command-line operations using npm, node, and a specific CLI tool to fetch auction data and manage court office codes. These commands are standard for the development and usage of this utility.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from an external source (courtauction.go.kr). While the source is a government entity, the skill implements a mandatory framing policy requiring the agent to warn users to verify information against the original court documents, providing a basic boundary for data integrity.
  • Ingestion points: Data is ingested via XHR endpoints and web scraping fallback from courtauction.go.kr.
  • Boundary markers: Explicit instructions require a disclaimer stating data is for reference only and requires manual verification.
  • Capability inventory: The skill provides read-only capabilities for searching notices and retrieving case details. No file-write or privileged system access was identified.
  • Sanitization: The skill processes raw site responses into structured JSON format, reducing the risk of direct instruction execution from the source data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 19, 2026, 01:41 PM
Security Audit — agent-trust-hub — data-court-auction