data-building-ledger

Pass

Audited by Gen Agent Trust Hub on Aug 19, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes a remote Model Context Protocol (MCP) server hosted at https://archhub-mcp.fly.dev/mcp. This server facilitates access to South Korean government building data without requiring individual API keys from the user. The service is hosted on Fly.io, a well-known application hosting platform.
  • [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection because it ingests and processes large volumes of building ledger and permit data from external sources.
  • Ingestion points: Data retrieved from the archhub MCP tools, including building_profile, building_ledger, building_floors, and price_history in SKILL.md.
  • Boundary markers: There are no explicit delimiters or specific instructions for the agent to ignore instructions embedded within the retrieved architectural data records.
  • Capability inventory: The skill provides read-only access to building statistics, history, and metadata; it does not contain tools for writing to the file system or executing arbitrary shell commands.
  • Sanitization: The skill relies on the veracity of data from the official government portal (data.go.kr) and uses [NOT_FOUND] markers to prevent hallucination, but does not implement content filtering for the data results.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 19, 2026, 01:41 PM
Security Audit — agent-trust-hub — data-building-ledger