ha-dashboard-create

Pass

Audited by Gen Agent Trust Hub on Feb 24, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection due to its interaction with external Home Assistant data.
  • Ingestion points: The skill retrieves state information and entity IDs from a Home Assistant instance using the get_states WebSocket command in scripts/create_dashboard.py and references/entity-patterns.md.
  • Boundary markers: The script does not utilize specific delimiters or instructions to the LLM to ignore potentially malicious instructions embedded within entity names or attributes.
  • Capability inventory: The skill has the capability to write to the Home Assistant configuration via the lovelace/dashboards/create and lovelace/config/save operations, as well as accessing system logs via system_log/list.
  • Sanitization: There is no evidence of sanitization or validation of the content of entity IDs or state data before they are processed or used to generate dashboard configurations.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 24, 2026, 05:23 PM