monitoring-capture-service

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

• Third-party content exposure detected (medium risk: 0.65). High: the skill’s runtime workflow uses Grafana MCP tools (e.g., search_dashboards, get_dashboard_by_uid, get_dashboard_panel_queries, query_prometheus, query_loki_logs) that ingest dashboard/panel query text and log contents returned by Grafana/Loki; those returned texts can include outsider-authored free text (e.g., dashboard/panel descriptions or log messages) which the agent may place into the LLM context.