Home Assistant YAML

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The documentation includes references to download legitimate community tools and services such as the Home Assistant Community Store (HACS) from https://get.hacs.xyz and the Ollama local LLM engine from https://ollama.com/install.sh. Both are well-established entities in the home automation and AI communities.
  • [REMOTE_CODE_EXECUTION]: Reference files like references/hacs-popular.md and references/integrations-ai-llm.md provide instructional examples of how to install software using piped shell commands (wget | bash and curl | sh). These are provided as guidance for manual user setup of trusted software and are not executed by the agent itself.
  • [COMMAND_EXECUTION]: Technical documentation across several files (references/backup-restore-migration.md, references/system-monitor.md, references/troubleshooting.md) lists common shell commands for troubleshooting, backup management (cp, tar, rclone), and system monitoring (ip, ping, ha core check). These examples are purely educational.
  • [SAFE]: Extensive analysis of the 90 provided files found no evidence of prompt injection, data harvesting, or hidden malicious logic. The skill includes strong instructions for the agent to use modern syntax and attribution, while proactively warning users against hardcoding credentials by utilizing the !secret feature.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 08:51 AM
Security Audit — agent-trust-hub — Home Assistant YAML