Control a device through capability operations like turn_on, set_brightness, set_target_temperature, or run_action.
AI agents invoke control_device to trigger actions in 米家 MCP Server. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
This tool triggers external physical operations on smart home devices (turning devices on/off, adjusting settings, running actions). These are real-world side effects on IoT hardware. While some actions may be reversible (e.g., turn_off after turn_on), the tool executes commands on physical systems, which fits Execute.
From the tool's definition 'Control a device through capability operations like turn_on, set_brightness, set_target_temperature, or run_action'
Documented attack patterns abuse exactly the kind of access control_device gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and 米家 MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for control_device:
{
"version": "1",
"default": "deny",
"tools": {
"control_device": {
"limits": [
{
"counter": "control_device_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}control_device stays usable, but rate-capped — a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Control a device through capability operations like turn_on, set_brightness, set_target_temperature, or run_action. It is categorised as a Execute tool in the 米家 MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the 米家 MCP Server MCP server in PolicyLayer and add a rule for control_device: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches 米家 MCP Server. Nothing to install.
control_device is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the control_device rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for control_device. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
control_device is provided by the 米家 MCP Server MCP server (javen-yan/miot-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from 米家 MCP Server, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
26 米家 MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.