observability-testing-patterns
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Node.js scripts located in the .claude/skills/qe-browser/ directory to perform dashboard assertions and visual diffing tasks (SKILL.md).
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it processes data from external monitoring sources. 1. Ingestion points: Data is ingested from Elasticsearch indices, Kibana saved objects, Jaeger traces, and web pages via esClient, kibanaApi, jaegerClient, and page.locator (SKILL.md). 2. Boundary markers: No explicit boundary markers or instructions to ignore embedded content were found in the provided patterns. 3. Capability inventory: The agent can execute local scripts, perform database queries, and make HTTP requests based on the results of its analysis (SKILL.md). 4. Sanitization: The provided snippets do not show explicit sanitization or filtering of the ingested external content before it is used in logic or assertions.
Audit Metadata