agent-observability-experiment-py-bootstrap

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

• Potentially malicious external URL detected (high risk: 0.70). The skill explicitly instructs at runtime to WebFetch Datadog docs (e.g. https://docs.datadoghq.com/llm_observability/) to answer feature questions and cite the page, which means external page content is fetched and injected into the agent's context and can directly influence prompts/output.