ha-integration-dev
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill is a technical reference and template library for Home Assistant developers. Analysis shows behavior consistent with the stated purpose.
- [PROMPT_INJECTION]: The conversation agent template in templates/conversation-agent/conversation_agent.py processes device states and executes actions based on LLM output. This pattern is susceptible to indirect prompt injection if device data contains malicious instructions. This is an intended design for such integrations and is mitigated via system prompt instructions.
- [COMMAND_EXECUTION]: The documentation in references/security.md contains snippets of unsafe code to demonstrate vulnerabilities like shell injection. These are explicitly labeled as insecure and provided for educational training purposes.
Audit Metadata