langsmith-observability

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

• Potentially malicious external URL detected (high risk: 0.90). The skill demonstrates pulling prompts from the LangSmith hub at runtime (client.pull_prompt("...")), which fetches remote prompt templates via the LangSmith API endpoint (e.g., https://api.smith.langchain.com) and therefore can directly control agent instructions.