sgds-layouts

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The SKILL.md workflow explicitly instructs the agent to fetch raw story files from public GitHub raw URLs (e.g., https://raw.githubusercontent.com/GovTechSG/sgds-web-component/... ) and to read/interpret that untrusted third-party content as part of adapting layouts, which could allow indirect prompt-injection via fetched content.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs agents at runtime to fetch raw story files (e.g. https://raw.githubusercontent.com/GovTechSG/sgds-web-component/master/stories/layouts/FullWidth/Default.stories.js) via curl and extract their HTML into the agent context, which means remote content can directly control the agent's output/instructions.