HA Gatekeeper FAQs

Question 1

What is HA Gatekeeper and what problem does it solve?

Accepted Answer

HA Gatekeeper is a reliability-first MCP server for Home Assistant, acting as a policy engine for AI-generated commands. It solves the critical problem of untrusted LLM tool calls by adding robust authorization, auditing, and validation layers before execution, preventing errors or unsafe actions.

Question 2

What are the prerequisites for running HA Gatekeeper?

Accepted Answer

To run HA Gatekeeper, you need Node.js >= 20, a running Home Assistant instance, and a long-lived access token from Home Assistant. Docker is the recommended and easiest way to deploy it.

Question 3

How does HA Gatekeeper ensure the safety and reliability of AI commands in Home Assistant?

Accepted Answer

It employs a comprehensive set of features including a four-tier policy engine, per-entity rate limiting, range clamping for values, fuzzy entity validation, and a three-state circuit breaker to protect against cascading failures and isolate degradations, ensuring dependable operation.

Question 4

What is the Four-Tier Policy Engine and how does it work?

Accepted Answer

The Four-Tier Policy Engine classifies AI commands into Unrestricted, Logged, Confirmed (requiring human approval), or Prohibited tiers. This allows granular control over which actions an AI can take, with entity-level overrides for tailored security on specific devices.

Question 5

Does HA Gatekeeper provide an audit trail for AI actions?

Accepted Answer

Yes, it features a tamper-evident audit log powered by SQLite with SHA-256 hash-chained entries for every policy decision. This ensures complete transparency and traceability of all AI-driven interactions within your Home Assistant setup.

HA Gatekeeper

HA Gatekeeper

Características Principales

Casos de Uso

Características Principales

Casos de Uso