use-design-md

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches design metadata and tokens from the vendor's catalog site and GitHub repository. Evidence includes curl -s https://getdesign.kr/llms.txt and curl -s https://raw.githubusercontent.com/CaesiumY/ko-design-md/main/services/<slug>.tokens.json in SKILL.md.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests unauthenticated markdown prose from an external web service and uses it as a brief to generate UI code. This creates a surface where malicious instructions embedded in the design guidelines could influence the agent's actions in the local repository. Ingestion points: https://getdesign.kr/services/<slug>/llms.txt (referenced in SKILL.md). Boundary markers: Absent. The agent is instructed to read the markdown content fully to capture design intent without specific delimiters or warnings to ignore embedded instructions. Capability inventory: Modifying local project files including Tailwind configurations, CSS files, and component source code (JSX/TSX). Sanitization: Absent. No mention of validation or filtering of the external markdown content before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 12:42 AM
Security Audit — agent-trust-hub — use-design-md