chatgpt-app-builder

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

• Third-party content exposure detected (high risk: 0.80). The skill's required workflow (e.g., fetch-and-render-data.md plus CSP and open-external-links guidance) instructs views/tools to fetch data from external APIs and open arbitrary external URLs (connectDomains/redirectDomains) and to return that structuredContent/content for the LLM to read and act on (e.g., booking/checkout), so untrusted third-party content can influence agent decisions.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)