HomeMCP Gateway
Provides a policy-governed gateway for Home Assistant tools, applying per-entity allow/deny rules to control lights and switches, read sensors, and deny locks and alarms, with dry-run defaults and an audit log of decisions. Connects to a live Home Assistant instance via URL and long-lived access token, or runs against a stub client.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
HomeMCP Gateway
HomeMCP Gateway is a prototype policy layer for Home Assistant tools. It applies per-entity allow/deny rules, defaults to dry-run, and records decisions in an audit log. It runs with a stub Home Assistant client by default and can use the separately maintained hass-ts SDK for live connections.
Run locally
Requirements: Node.js and pnpm.
pnpm install
pnpm devOpen http://localhost:3000/demo for the interactive demo. pnpm demo runs the command-line walkthrough. Stub mode requires no credentials.
Related MCP server: MCP Gatekeeper
Live Home Assistant
Set HASS_URL and HASS_TOKEN in the environment to connect to an instance. The token is a Home Assistant long-lived access token; keep it private and do not commit it. Live mode can control devices when dry-run is disabled. Review the policy and every requested operation before enabling side effects.
API
Method | Path | Purpose |
GET |
| List available MCP tools |
POST |
| Request a tool call subject to policy |
GET / PUT |
| Read or update entity permissions |
GET |
| View recent decisions |
GET |
| Check service mode and status |
The default policy allows selected light/switch operations, permits sensor reads, and denies locks and alarms. Dry-run is enabled by default.
Build and checks
pnpm typecheck
pnpm test
pnpm buildSecurity limitations
This is a prototype, not a security-certified gateway. Do not expose it directly to the public internet: the API has no built-in user authentication. Put it behind a trusted network boundary and authenticated proxy. The audit/policy store is local JSON and is not designed for concurrent multi-instance deployments. Review policies carefully; live calls can operate real devices.
Docker deployment is available through a manually triggered workflow after provider settings are configured.
License
Licensed under the PolyForm Noncommercial License 1.0.0. This source-available license restricts commercial use; it is not an OSI-approved open-source license. See the license for terms and exceptions. Third-party dependencies retain their own licenses.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Runtime permission, approval, and audit layer for AI agent tool execution.
Deterministic allow/require_approval/deny verdicts for agent actions, before they happen.
Authorize AI agent tool calls before they run: policy-as-code, human approval, signed receipts.
Fail-closed policy guardrails for AI agents running kubectl, terraform, helm, and argocd.
Related MCP Servers
- AlicenseAqualityCmaintenanceLocal zero-trust permission gateway for AI agents. Enforces policy-based tool authorization, human approvals, scoped permissions, and cryptographically verifiable audit logs.4128 PyPI5Apache 2.0
- FlicenseNot gradedqualityDmaintenanceEnables policy-enforced access to dangerous tools like file read/write/delete and shell execution, with approval workflows, risk classification, and audit logging.-
- FlicenseNot gradedqualityCmaintenanceEnables controlled AI-agent access to enterprise-shaped tools with a deny-by-default gated write path, human approval, dry-run execution, and append-only audit logging.1-
- AlicenseNot gradedqualityCmaintenanceProvides permission gates and tamper-evident audit logging for AI agent tool executions, with declarative policies, consent ladders, and hash-chained verification.MIT