Moltbot Home Assistant

**Judgment:** Solid, capable Home Assistant control skill with a well‑designed CLI and thoughtful safety mechanisms. Appropriate for serious smart home control, including actions that affect the physical environment, as long as safety settings and tokens are managed carefully. **Strengths & Functionality** - Covers core Home Assistant operations: entity discovery (`list`), state inspection (`state`), simple control (`on`, `off`, `toggle`), attribute setting (`set`), and generic service invocation (`call`, including JSON payloads). - Works across multiple domains (lights, switches, covers, climate, scenes, automations, scripts), making it broadly useful for most Home Assistant setups. - JSON output support (`--json`) is suitable for structured parsing by agents. **Safety & Risk Profile** - Provides a configurable safety system with levels 0–3 and a default of **level 3**, which requires explicit confirmation for critical domains (`lock`, `alarm_control_panel`, `cover`). - Enforces a clear confirmation workflow for critical operations (e.g., opening garage doors, manipulating locks/alarms) and explicitly warns agents to never use `--force` without user consent. - Supports `blocked_entities` and `allowed_entities` (with wildcard support) to hard‑restrict what the tool can touch, plus logging of all write actions for audit. - Residual risks: - Misconfiguration (e.g., lowering safety level, failing to block high‑risk entities) can expose powerful operations without adequate friction. - The Home Assistant long‑lived access token grants broad control; compromise of that token or the host running `moltbot-ha` could expose the home. - Safety relies partly on the agent respecting the `--force` rule; misbehaving or poorly configured agents could still perform undesirable critical actions if safeties are relaxed. **Recommended Scenarios** - **Good fit:** - Users with an existing Home Assistant deployment who want conversational or agent‑driven control of lights, switches, scenes, climate, and automations. - Setups where the operator can configure `allowed_entities`/`blocked_entities` to tightly scope what the agent can control. - Use cases needing structured data about sensors and entity states for higher‑level reasoning or decision‑making. - **Use with extra caution or avoid:** - Environments where the Home Assistant instance controls highly sensitive systems (door locks, alarms, main power, safety‑critical devices) without properly configured safety levels and allow/block lists. - Multi‑user or untrusted‑input scenarios if you cannot guarantee correct handling of confirmations and `--force` usage. **Bottom line:** A robust and versatile Home Assistant integration that is well‑suited for agent use, provided that you carefully configure safety settings and access scopes before exposing it to end users or automated workflows.