observability-llm-obs

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill links to official documentation from Elastic (elastic.co) and OpenTelemetry (opentelemetry.io). These are well-known and trusted technology sources.
  • [PROMPT_INJECTION]: The skill is designed to summarize LLM trace and log data, which inherently involves processing untrusted external inputs.
  • Ingestion points: Elasticsearch data streams (traces*, logs*, metrics*) containing LLM interaction data such as prompt and response messages.
  • Boundary markers: None specified. The skill does not define specific delimiters or instructions for separating trace data from agent instructions.
  • Capability inventory: Ability to execute ES|QL queries and interact with Elasticsearch and Kibana APIs for Alerting and SLO management.
  • Sanitization: No explicit filtering or sanitization of ingested content is mandated within the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 04:51 PM