Token导航 LogoToken导航

Agent Skill · 研究检索

home-assistant(家庭助理)

home-assistant 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

home-assistant 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为接触敏感数据;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 42,917。

总安装 1,028周安装 42GitHub Stars 125下载 329核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:home-assistant(家庭助理)
来源仓库:https://github.com/sammcj/agentic-coding
仓库路径:skills/home-assistant
安装命令:
npx skills add https://github.com/sammcj/agentic-coding --skill home-assistant
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/sammcj/agentic-coding --skill home-assistant

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

home-assistant 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。它支持按关键词、任务类型或来源仓库进行信息聚合与过滤,帮助 Agent 快速缩小范围。

通过 npx skills add 命令从指定 GitHub 仓库安装,具体路径为 skills/home-assistant。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/home-assistant

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/sammcj/agentic-coding --skill home-assistant;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

home-assistant 怎么安装?

在终端执行 npx skills add https://github.com/sammcj/agentic-coding --skill home-assistant。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

home-assistant 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

home-assistant 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

home-assistant 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

home-assistant 用的人多吗?

公开数据是累计安装 1,028 次,近一周安装 42 次,下载 329 次,GitHub 125 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

home-assistant 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
home-assistant

家庭助理 · 本页

1,028125接触敏感数据未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Home Assistant Helper

Overview

This skill provides tools and workflows for working with Home Assistant installations. It enables querying the live HA instance for entities, services, and configuration data, debugging automations, and generating YAML configurations for automations and dashboards.

Key Capabilities:

  • Query entities, states, and services from the live HA installation
  • Search for similar entities to use as examples
  • Check automation states and execution history
  • Generate YAML configurations for copy/paste into HA
  • Find real examples from the user's setup to inform new configurations

Core Workflow

When helping with Home Assistant tasks, follow this general approach:

  1. Understand the requirement - What is the user trying to accomplish?
  2. Discover existing entities - Use scripts to find relevant entities in their setup
  3. Find similar examples - Search for existing automations or entities that do something similar
  4. Generate YAML - Create well-formed YAML that can be copied directly into HA
  5. Explain the configuration - Describe what the YAML does and how to install it

Available Scripts

All scripts require the HA_TOKEN environment variable to be set, which contains the Home Assistant long-lived access token. The HA instance is available at http://homeassistant.local:8123.

Important: All scripts now use uv with inline PEP 723 dependency declarations and the homeassistant-api library for consistent, maintainable code. Dependencies are automatically installed by uv on first run.

Entity Discovery

ha_get_entities.py [domain]

Retrieve all entities, optionally filtered by domain.

Usage:

uv run scripts/ha_get_entities.py           # All entities
uv run scripts/ha_get_entities.py light     # Just lights
uv run scripts/ha_get_entities.py sensor    # Just sensors

When to use: To discover what entities are available, especially when building new automations.

ha_get_state.py <entity_id>

Get the current state and attributes of a specific entity.

Usage:

uv run scripts/ha_get_state.py light.living_room

When to use: To check current state, available attributes, or confirm an entity exists.

ha_search_similar_entities.py <pattern>

Search for entities matching a pattern in their entity_id or friendly_name.

Usage:

uv run scripts/ha_search_similar_entities.py "bedroom"
uv run scripts/ha_search_similar_entities.py "motion"
uv run scripts/ha_search_similar_entities.py "temperature"

When to use: To find entities related to what the user wants to automate. This is especially useful for finding examples before creating new automations.

Automation Management

ha_get_automations.py [search_term]

Retrieve all automations, optionally filtered by search term.

Usage:

uv run scripts/ha_get_automations.py              # All automations
uv run scripts/ha_get_automations.py motion       # Automations with 'motion'
uv run scripts/ha_get_automations.py light        # Automations with 'light'

When to use: To find existing automations that are similar to what the user wants to create. Use these as templates.

Automation Trace Analysis

ha_list_traces.py [automation_id]

List automation execution traces with timestamps and status.

Usage:

uv run scripts/ha_list_traces.py                    # All traces
uv run scripts/ha_list_traces.py 1761430536701      # Specific automation traces (using numeric ID)

Important: Use the numeric automation ID (e.g., 1761430536701), not the full entity_id (e.g., automation.bedroom_light). You can find the numeric ID in the automation's attributes from ha_get_automations.py output.

When to use: To see recent automation runs, their status, and timing. Use this to identify which run IDs to investigate further.

Output includes:

  • Run ID (for use with ha_get_trace.py)
  • Timestamp
  • Execution state (stopped, running, etc.)
  • Script execution status (finished, failed_single, failed_conditions, etc.)
  • Last step executed
  • Any errors
ha_get_trace.py <automation_id> <run_id>

Get detailed step-by-step trace for a specific automation run.

Usage:

uv run scripts/ha_get_trace.py 1761430536701 1ceef6b2b6f63a8745eb5dba3fe12f71

Important: Use the numeric automation ID (e.g., 1761430536701), not the full entity_id.

When to use: To debug a specific automation run. Shows the complete execution path including:

  • Trigger details
  • Condition evaluations (pass/fail)
  • Actions executed
  • Variables at each step
  • Timing information
  • Error details if failed

Tip: Get the run_id from ha_list_traces.py output.

ha_trace_summary.py <automation_id>

Get aggregated statistics for an automation's execution history.

Usage:

uv run scripts/ha_trace_summary.py 1761430536701

Important: Use the numeric automation ID (e.g., 1761430536701), not the full entity_id.

When to use: To understand automation reliability and performance over time.

Output includes:

  • Total runs
  • Success/failure counts and rates
  • Average/min/max execution times
  • Common error patterns
  • Distribution of where executions complete

Service Discovery

ha_get_services.py [domain]

Get all available services with descriptions and field information.

Usage:

uv run scripts/ha_get_services.py              # All services
uv run scripts/ha_get_services.py light        # Just light services
uv run scripts/ha_get_services.py climate      # Just climate services

When to use: To discover what services are available and what parameters they accept.

Configuration

ha_get_config.py

Get Home Assistant configuration including version, location, and components.

Usage:

uv run scripts/ha_get_config.py

When to use: To understand the HA setup, available integrations, or system information.

ha_get_config_entries.py [domain]

Get Home Assistant config entries, optionally filtered by domain. This is essential for services that require a config_entry_id, such as telegram_bot.send_message.

Usage:

uv run scripts/ha_get_config_entries.py              # All config entries
uv run scripts/ha_get_config_entries.py telegram_bot # Just Telegram bots
uv run scripts/ha_get_config_entries.py mqtt         # Just MQTT entries

When to use: When you need to get config_entry_id for services like Telegram notifications, or to discover what integrations are configured.

Service Calling

ha_call_service.py <domain> <service> <json_data>

Call a Home Assistant service (use with caution).

Usage:

uv run scripts/ha_call_service.py light turn_on '{"entity_id": "light.living_room"}'

When to use: Rarely. Generally only for testing or when the user explicitly asks to control something.

Typical Workflows

Creating a New Automation

  1. Understand the goal - Ask clarifying questions about triggers, conditions, and actions
  2. Find similar entities - Use ha_search_similar_entities.py to find relevant entities
  3. Search for similar automations - Use ha_get_automations.py with search terms to find examples
  4. Review existing automation - If a similar one exists, examine its structure
  5. Generate YAML - Create well-formatted YAML with:

- Descriptive alias - Clear description - Appropriate triggers - Relevant conditions - Necessary actions - Proper mode (single, restart, queued, parallel)

  1. Provide copy-paste YAML - Format for easy copying into HA configuration
  2. Explain - Describe what the automation does and how to add it to HA

Debugging an Automation

  1. Check automation status - Use uv run scripts/ha_get_state.py automation.automation_name to check:

- Current state (on/off) - Last triggered time - Current execution count - Automation mode

  1. Review execution history - Use trace analysis tools (use numeric automation ID from attributes):

- uv run scripts/ha_trace_summary.py <numeric_id> - Get success rate and common issues - uv run scripts/ha_list_traces.py <numeric_id> - See recent runs - uv run scripts/ha_get_trace.py <numeric_id> <run_id> - Detailed step-by-step trace

  1. Analyze trace data - Look for:

- Which trigger fired - Whether conditions passed or failed - Which actions executed successfully - Where the automation stopped or errored - Variable values at each step

  1. Check related entities - Use uv run scripts/ha_get_state.py to verify trigger entities are in expected states
  2. Review the automation configuration - Use uv run scripts/ha_get_automations.py to see the full automation details
  3. Identify the issue - Based on trace data, state information, and configuration review
  4. Suggest fix - Provide corrected YAML or configuration changes

Trace-based Debugging Workflow:

# Step 1: Get the automation details to find the numeric ID
uv run scripts/ha_get_automations.py problem_automation

# Step 2: Check if automation has been running and get stats (use numeric ID from step 1)
uv run scripts/ha_trace_summary.py 1761430536701

# Step 3: List recent runs to find failing ones (use numeric ID)
uv run scripts/ha_list_traces.py 1761430536701

# Step 4: Get detailed trace for a specific failed run (use numeric ID)
uv run scripts/ha_get_trace.py 1761430536701 <run_id_from_step_3>

# Step 5: Examine the trace to see exactly where and why it failed

Modifying a Dashboard

  1. Understand desired changes - What should the dashboard show?
  2. Find relevant entities - Use entity discovery scripts
  3. Generate Lovelace YAML - Create dashboard card configuration
  4. Provide copy-paste YAML - User will manually add to their dashboard
  5. Explain card configuration - Describe options and customization

Exploring Entity States

  1. Use search or domain filtering - Find entities of interest
  2. Check specific states - Get detailed state information
  3. Report findings - Present relevant information clearly

Sending Telegram Notifications

Telegram notifications require using the telegram_bot.send_message service with a config_entry_id parameter (not the notify service pattern).

Workflow:

  1. Get the Telegram bot config_entry_id - Use uv run scripts/ha_get_config_entries.py telegram_bot to find the config entry ID
  2. Use the telegram_bot.send_message service - Include the config_entry_id in the action data

Example Automation with Telegram Notification:

alias: Example Telegram Alert
description: Send a Telegram notification when something happens
triggers:
  - entity_id: binary_sensor.front_door
    to: "on"
    trigger: state
conditions: []
actions:
  - action: telegram_bot.send_message
    data:
      message: "Front door opened at {{ now().strftime('%I:%M %p') }}"
      config_entry_id: 01JZE11D7Y6B7C3WCARWVZRYNH  # Get this from ha_get_config_entries.py
mode: single

Note: The config_entry_id is specific to your Telegram bot configuration. Always use uv run scripts/ha_get_config_entries.py telegram_bot to get the correct ID for your setup.

Important Notes

YAML Output Format

When generating YAML configurations, always:

  • Use proper YAML formatting with 2-space indentation
  • Include helpful comments where appropriate
  • Provide descriptive aliases and descriptions
  • Use appropriate trigger platforms (state, time, numeric_state, etc.)
  • Include mode settings (single, restart, queued, parallel)
  • Format for easy copy-paste into HA

Manual Installation Required

The user must manually copy/paste generated YAML into Home Assistant. Make this clear and provide instructions:

  • For automations: Configuration → Automations & Scenes → Add Automation → Edit in YAML
  • For dashboards: Dashboard → Edit Dashboard → Raw Configuration Editor
  • For configuration.yaml additions: Edit the file and restart HA

Browser Automation for Screenshots

If the user asks for screenshots of dashboards or wants to see the current UI state, the Playwright browser automation tools can be used to navigate to the HA instance and capture screenshots. The user will need to handle authentication.

Service Calls

Be cautious about calling services that change state. Generally only do this when explicitly requested by the user or for testing purposes.

Entity Naming

All entities follow the pattern domain.object_id where common domains include:

  • light - Lights
  • switch - Switches
  • sensor - Sensors (read-only)
  • binary_sensor - Binary sensors (on/off)
  • climate - Thermostats
  • automation - Automations
  • script - Scripts
  • input_boolean, input_number, input_select, input_text, input_datetime - Helper entities
  • person - People
  • device_tracker - Device tracking
  • camera - Cameras
  • media_player - Media players
  • cover - Covers (blinds, garage doors)
  • fan - Fans
  • lock - Locks

Common Trigger Platforms

  • state - Entity state changes
  • numeric_state - Numeric value crosses threshold
  • sun - Sunrise/sunset
  • time - Specific time
  • time_pattern - Time pattern (every N minutes)
  • event - HA event fired
  • webhook - Webhook trigger
  • zone - Enter/leave zone
  • device - Device-specific trigger

Common Condition Platforms

  • state - Entity state equals value
  • numeric_state - Numeric comparison
  • time - Time window
  • sun - Before/after sunrise/sunset
  • zone - Person in zone
  • template - Template evaluation

Automation Modes

  • single - Don't start new run if already running
  • restart - Restart automation if triggered while running
  • queued - Queue runs if already running
  • parallel - Allow multiple simultaneous runs

平台分布

Codex

35.32%
按下载量换算116

Claude

30.05%
按下载量换算99

Cursor

20.17%
按下载量换算66

Gemini CLI

8.26%
按下载量换算27

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills