kibana-dashboards

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCREDENTIALS_UNSAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The documentation mentions a setup command curl -fsSL https://elastic.co/start-local | sh. This script is hosted on the official Elastic domain (elastic.co) and is a standard method for setting up local development environments for the author's software.
  • [CREDENTIALS_UNSAFE]: The tool handles authentication via environment variables (KIBANA_API_KEY, KIBANA_PASSWORD, KIBANA_CLOUD_ID). This is standard practice for CLI-based agent skills to interact with remote services.
  • [COMMAND_EXECUTION]: The script scripts/kibana-dashboards.js uses the native fetch API to perform CRUD operations on Kibana. It provides an option to disable TLS validation (KIBANA_INSECURE), which is a known security trade-off for local development.
  • [EXTERNAL_DOWNLOADS]: The skill references official documentation and GitHub repositories owned by Elastic. These are trusted sources for the specified author.
  • [PROMPT_INJECTION]: The skill has an indirect injection surface because it ingests JSON data from external files or stdin. Ingestion points: scripts/kibana-dashboards.js (loadSpec). Boundary markers: absent. Capability inventory: network requests (fetch) in scripts/kibana-dashboards.js. Sanitization: absent. This could allow maliciously crafted dashboard definitions to influence the agent, though its capabilities are restricted to the Kibana API.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 04:50 PM