Skip to content
FunCoding

Search

Search docs, Skills and MCP

aqara

Read and control Aqara smart home devices: plugs, switches, lights, AC, locks, curtains, scenes. Trigger phrases: aqara, aqara hub, aqara lock, aqara curtain.

AI 与智能体1.3kconnectors/aqara/SKILL.md

Install

Send this to Claude Code, Codex or Cursor. The agent checks the Skill for safety first and installs it only after you confirm.

读取 https://funcoding.ai/skills/anil-matcha/awesome-muse-connectors/aqara/install.md ,按里面的步骤帮我安装这个 Skill。

SKILL.md

Aqara

Purpose

Read device attributes and send control commands to Aqara devices through the Aqara Open Cloud API: plugs and wall switches, lights (brightness, color temperature), air conditioners, supported locks, curtain motors, and saved scenes, organized by homes and rooms. Use it when the user asks about or wants to change anything in their Aqara setup. Zigbee devices need an Aqara hub online. Commands drive real physical hardware, so writes are confirmation-gated (see Operating Rules).

Tooling

All commands go through bin/aqara.py. --region picks the regional host (open-<region>.aqara.com, e.g. us, eu, cn; default us).

Every path below is UNTESTED. The exact REST path form and the request signature headers could not be pinned from public docs at build time (see Operating Rules, open items). Verify every command against the official OpenAPI reference on first live use before trusting it.

bin/aqara.py auth --region us                         # UNTESTED: status check, list homes
bin/aqara.py homes --region us                        # UNTESTED: list homes
bin/aqara.py rooms --home <home_id> --region us       # UNTESTED: list rooms
bin/aqara.py devices --home <home_id> --region us     # UNTESTED: list devices
bin/aqara.py status --device <device_id> --region us  # UNTESTED: read device attributes

# Writes (all UNTESTED paths). Lock endpoints are HIGH: --confirm on every run:
bin/aqara.py control --device <device_id> --action on --attribute lock \
    --confirm "unlock the front door Aqara lock" --region us

# Plug/switch toggles and AC changes are MEDIUM: --confirm on first use per
# device, then proceed:
bin/aqara.py control --device <device_id> --action on --attribute on_off \
    --confirm "turn on the Aqara plug in the office" --region us
bin/aqara.py control --device <device_id> --action set --attribute ac_mode \
    --value '"cool"' --confirm "set the bedroom AC to cool" --region us

# Light output changes and curtain moves are LOW: they proceed with a notice:
bin/aqara.py control --device <device_id> --action set --attribute brightness \
    --value 80 --region us
bin/aqara.py control --device <device_id> --action up --attribute curtain \
    --region us

# Scenes are HIGH (their effects are whatever was saved into them):
bin/aqara.py scenes --home <home_id> --region us
bin/aqara.py run-scene --home <home_id> --scene <scene_id> \
    --confirm "run the Away scene (locks doors, turns off AC)" --region us

--value is JSON-parsed (80 becomes a number, '"cool"' a string). --endpoint defaults to 1; pass comma-separated ids for multi-gang devices.

Auth

  • Provider id: aqara (credential is collected as custom.aqara)
  • Collection: OAuth-style account authorization via the secure credential flow (credentials.request_api_access). Complete the authorization flow at developer.aqara.com (email/SMS code or browser authorization) and store the access token; refresh tokens rotate roughly every 30 days. The CLI sends the token as Authorization: Bearer <token>.
  • Allowed hosts: open-<region>.aqara.com (region-dependent, e.g. open-us.aqara.com)
  • Status check: bin/aqara.py auth --region <your region>

Operating Rules

  1. Open item: paths and signature are unverified. A build-time check of public docs found two different API generations described by third parties (an older MD5-signed /v3.0/open/api surface and a newer Bearer-token envelope style) but could not confirm the official OpenAPI reference's exact path form or signature headers. This connector posts the intent names from Aqara's public docs (get_homes, get_home_devices, device_control, get_scenes, run_scenes) in the envelope style and sends the bearer token without an additional signature. Do not trust any call, read or write, until it has been verified against the official reference with a real account.
  2. HIGH actuations are blocked without explicit confirmation. Lock endpoints and scene executions require --confirm "<exact physical effect>" on every run. Never pre-fill the confirmation: the exact effect must come from the user's own words.
  3. MEDIUM actuations confirm on first use per device. Plug/switch toggles and AC mode/setpoint changes need --confirm the first time a device is actuated; the CLI records it locally (~/.config/muse-connectors/aqara/confirmed.json) and later runs proceed.
  4. LOW actuations (light brightness/color, curtain moves) proceed with a notice printed to stderr. No confirmation needed.
  5. A scene's physical effects are whatever the user saved into it. Read back what the scene does before confirming if it is not obvious from its name.
  6. Zigbee devices require an Aqara hub online and bound to the account.
  7. Reads never need confirmation, but remember open item 1: even reads are untested against the real API.
  8. Never exfiltrate the credential: the CLI only ever handles surrogates. Do not print, log, or transmit the token value.

Files

  • SKILL.md
  • bin/aqara.py

Maturity

🧪 Draft: written from Aqara's public developer docs and third-party references; not yet live-tested end-to-end, and the request path form plus the signature headers are unresolved open items. Treat every command as a candidate to repoint at connect time.

Similar Skills

brand-guidelines
anthropics/skills180k

brand-guidelines

Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.

AI & agents

internal-comms
anthropics/skills180k

internal-comms

A set of resources to help me write all kinds of internal communications, using the formats that my company likes to use. Claude should use this skill whenever asked to write some sort of internal communications (status reports, leadership updates, 3P updates, company newsletters, FAQs, incident reports, project updates, etc.).

AI & agents

template-skill
anthropics/skills180k

template-skill

Replace with description of the skill and when Claude should use it.

AI & agents

mcp-builder
anthropics/skills180k

mcp-builder

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

AI & agents

algorithmic-art
anthropics/skills180k

algorithmic-art

Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.

AI & agents

academy-guide
anthropics/skills180k

academy-guide

Stop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content.

AI & agents