chatgpt-apps
Warn
Audited by Snyk on Mar 25, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.80). The skill invokes the OpenAI developer docs MCP at runtime (e.g. https://developers.openai.com/mcp and the referenced pages like https://developers.openai.com/apps-sdk/build/mcp-server/) as a required “docs-first” dependency whose fetched content is used to determine scaffolding prompts and generated code, so the external content can directly control agent instructions.
Issues (1)
W012
MEDIUMUnverifiable external dependency detected (runtime URL that controls agent).
Audit Metadata