Token导航 LogoToken导航

Agent Skill · 研究检索

weather(天气)

weather 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。

weather 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,权限为需要联网;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 73,674 个 Skill 里排第 69,087。

总安装 220周安装 9GitHub Stars 公开资料未说明下载 71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:weather(天气)
来源仓库:https://github.com/mateeb11/weather-forecast
仓库路径:skills/weather
安装命令:
npx skills add https://github.com/mateeb11/weather-forecast --skill weather
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/mateeb11/weather-forecast --skill weather

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。

weather 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。适用于需要根据关键词或任务场景进行信息定位的研究与检索任务。通过 npx skills add 命令从 GitHub 仓库安装并使用该技能。

安装前需确认权限范围、维护状态及是否触发联网或文件操作。建议结合原始 README 和仓库内容进一步核验具体用法。

来源仓库内路径:skills/weather

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/mateeb11/weather-forecast --skill weather;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

    它需要需要联网,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

weather 怎么安装?

在终端执行 npx skills add https://github.com/mateeb11/weather-forecast --skill weather。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

weather 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。

weather 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

weather 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。

weather 用的人多吗?

公开数据是累计安装 220 次,近一周安装 9 次,下载 71 次。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。

weather 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

  • 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

  • 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
weather

天气 · 本页

220—需要联网未明确通过

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

Weather Skill

Provides current weather and forecasts for any location using Open-Meteo free APIs called directly via WebFetch:

  • Open-Meteo Geocoding for resolving location names to coordinates
  • Open-Meteo Forecast for weather data

Workflow

  1. Understand the user's request. Extract the location, time range, and whether they're planning travel. See examples below.
  2. Geocode the location using WebFetch to call the Nominatim API (see below). If the result is ambiguous, ask the user to clarify.
  3. Fetch weather using WebFetch to call the Open-Meteo API with the lat/lon and appropriate parameters (see below).
  4. Respond to the user with a clear summary. See example outputs below.
  5. If travel intent is detected, also read references/clothing_guide.md and include relevant clothing advice.

API Calls

Step 1: Geocode with Open-Meteo Geocoding

Use WebFetch to GET:

https://geocoding-api.open-meteo.com/v1/search?name={location}&count=3&language=en&format=json

Replace {location} with the URL-encoded location string (e.g., Tokyo, Riyadh).

Response is a JSON object with a results array. Each item has:

  • name — location name
  • latitude, longitude — coordinates (floats)
  • country — country name
  • admin1 — state/region (optional)
  • population — population count (optional, useful for disambiguation)

Ambiguity logic: If there are ≥ 2 results in different countries or regions and neither clearly dominates by population, treat the result as ambiguous. Present each candidate's name, admin1, and country and ask the user which they meant.

If the results key is missing or the array is empty, the location was not found — ask the user to be more specific.

Step 2: Fetch Weather from Open-Meteo

Use WebFetch to GET the Open-Meteo forecast URL, building it from the resolved coordinates and the desired resolution.

Base URL:

https://api.open-meteo.com/v1/forecast?latitude={lat}&longitude={lon}&timezone=auto&forecast_days=16

Always include the current parameter:

&current=temperature_2m,apparent_temperature,weather_code,wind_speed_10m,relative_humidity_2m,precipitation

Add parameters based on resolution:

ResolutionAdditional params
current (now + next 24h)&hourly=temperature_2m,apparent_temperature,precipitation_probability,precipitation,weather_code,wind_speed_10m
hourly (specific day / short range)&hourly=temperature_2m,apparent_temperature,precipitation_probability,precipitation,weather_code,wind_speed_10m&daily=temperature_2m_max,temperature_2m_min,apparent_temperature_max,apparent_temperature_min,precipitation_sum,precipitation_probability_max,weather_code,wind_speed_10m_max
daily (multi-day / weekly)&daily=temperature_2m_max,temperature_2m_min,apparent_temperature_max,apparent_temperature_min,precipitation_sum,precipitation_probability_max,weather_code,wind_speed_10m_max

Optional params:

  • &past_days=N — include N historical days (0–92). Use only when the user asks about past weather.
  • &timezone=Asia/Tokyo — override auto timezone (rarely needed).

Response structure:

  • current object: time, temperature_2m, apparent_temperature, weather_code, wind_speed_10m, relative_humidity_2m, precipitation
  • hourly object: time[], temperature_2m[], apparent_temperature[], precipitation_probability[], precipitation[], weather_code[], wind_speed_10m[]
  • daily object: time[], temperature_2m_max[], temperature_2m_min[], apparent_temperature_max[], apparent_temperature_min[], precipitation_sum[], precipitation_probability_max[], weather_code[], wind_speed_10m_max[]

WMO Weather Code Table

Use this table to convert the numeric weather_code in API responses to human-readable descriptions:

CodeDescription
0Clear sky
1Mainly clear
2Partly cloudy
3Overcast
45Foggy
48Rime fog
51Light drizzle
53Moderate drizzle
55Dense drizzle
56Light freezing drizzle
57Dense freezing drizzle
61Slight rain
63Moderate rain
65Heavy rain
66Light freezing rain
67Heavy freezing rain
71Slight snow
73Moderate snow
75Heavy snow
77Snow grains
80Slight showers
81Moderate showers
82Violent showers
85Slight snow showers
86Heavy snow showers
95Thunderstorm
96Thunderstorm w/ hail
99Thunderstorm w/ heavy hail

How to Interpret User Requests

The agent should interpret the user's natural language to determine three things:

1. Location

Extract the place name from the query. Examples:

  • "Weather in Jeddah now" → "Jeddah"
  • "Forecast for London this weekend" → "London"
  • "I'm going to Tokyo next week" → "Tokyo"
  • "How windy will it be in AlUla on Friday?" → "AlUla"

If no location is mentioned, ask the user.

2. Time Range, Resolution, and Past Weather

User saysResolutionpast_daysNotes
"now" / "current" / no time specifiedcurrent (also include hourly for next 24h)0Forecast request always uses 16 days.
"today" / "tonight" / "tomorrow" / "tomorrow morning"hourly0Use hourly data to answer specific time windows.
"this weekend" / "on Friday" / "hourly for next 12 hours"hourly0Filter the returned series in the response.
"next 3 days"hourly0Summarize from hourly data.
"next 7 days" / "next week" / "first week of March"daily0Summarize relevant days from daily data.
"March" / "in April" / "05/01 to 06/30"daily0If outside forecast horizon, clearly state limits.
"past weather" / "last 3 days" / "yesterday"daily (or hourly for short windows)set from requestIf user says "past" without a number, default to 7.

The user may specify dates in various formats (MM/DD, DD/MM, "May 1st", "1 March", etc.) — interpret based on context and filter the returned timeline accordingly. Use the current year unless another year is stated or implied.

Use daily for ranges > 3 days, hourly for shorter windows, current for right-now queries (but still include hourly for context).

Default (no time specified): current conditions + next 24 hours hourly.

Remember: Forecast is fixed to 16 days. For past requests, use past_days (0..92).

3. Travel Intent

Detect if the user is planning to go somewhere. Look for keywords like: pack, wear, travel, visiting, going to, headed to, trip, vacation, holiday, outing, flight, should I bring

If travel intent is detected → also include clothing guidance (see below).


Example Outputs

Example 1: Simple current weather

User: "What's the weather in Jeddah now?"

Response:

Weather in Jeddah, Saudi Arabia Right now it's Clear sky with a temperature of 28°C (feels like 30°C). Wind is light at 12 km/h from the north, and humidity is at 45%. No precipitation expected. Next few hours: - 3 PM: 29°C, Clear sky, 0% rain - 4 PM: 28°C, Partly cloudy, 0% rain - 5 PM: 26°C, Clear sky, 0% rain *(and so on...)*

Example 2: Weekend forecast

User: "Forecast for London this weekend"

Response:

Weekend Forecast for London, UK Saturday, Mar 1: Highs of 11°C, lows of 5°C. Moderate rain expected (70% probability, ~4mm). Wind up to 25 km/h. Sunday, Mar 2: Highs of 13°C, lows of 6°C. Partly cloudy, 20% rain chance. Lighter wind at 15 km/h. Overall: A wet Saturday clearing into a drier Sunday. Bring an umbrella for Saturday.

Example 3: Travel planning with clothing advice

User: "I'm going to Tokyo next week — what should I pack?"

Response:

Tokyo, Japan — Next Week Forecast (Mar 3–9) | Day | High | Low | Conditions | Rain | Wind | | --- | --- | --- | --- | --- | --- | | Mon | 12°C | 4°C | Partly cloudy | 10% | 15 km/h | | Tue | 10°C | 3°C | Slight rain | 65% | 20 km/h | | ... | ... | ... | ... | ... | ... | What to Wear: Expect cold conditions (lows near 3°C) with rain mid-week. Pack: - Warm layers: fleece or wool sweater + insulated jacket - Waterproof jacket and umbrella for Tuesday/Wednesday - Closed-toe waterproof shoes - Scarf and light gloves for mornings/evenings *See: clothing_guide.md — sections: cold, rainy*

Example 4: Specific condition query

User: "How windy will it be in AlUla on Friday?"

Response:

Wind Forecast for AlUla, Saudi Arabia — Friday, Feb 28 Wind speeds will peak at 35 km/h in the early afternoon, settling to around 15 km/h by evening. Morning winds are lighter at ~10 km/h. ⚠️ Moderately windy conditions expected midday — secure loose items and consider a windbreaker if you'll be outdoors.

Warnings

Include warnings when conditions are notable:

  • Extreme heat: temp ≥ 40°C
  • Extreme cold: temp ≤ -10°C
  • Strong wind: ≥ 50 km/h
  • Heavy rain: ≥ 20mm precipitation

Clothing Guidance

When travel intent is detected, read references/clothing_guide.md and include relevant advice in the response.


Error Handling

  • Location not found: The Nominatim response is an empty array. Ask the user to be more specific (add city, country).
  • Ambiguous location: Present the alternatives and ask which one.
  • WebFetch failure: If WebFetch returns an error or non-JSON response, inform the user that the weather service is temporarily unavailable and suggest retrying.
  • Beyond forecast horizon: Inform the user of the 16-day limit, return what's available.
  • No time specified: Default to current + next 24 hours.

平台分布

Codex

36.04%
按下载量换算26

Claude

32.4%
按下载量换算23

Cursor

18.4%
按下载量换算13

Gemini CLI

8.33%
按下载量换算6

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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