langgraph-adk

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires querying context7 at scaffold/runtime (e.g. /websites/langchain_oss_javascript_langgraph) to verify current LangChain/LangGraph APIs and uses that fetched API documentation to drive generated agent code, so the external content directly controls the prompts/instructions the skill emits and is a required runtime dependency.