Free SKILL.md scraped from GitHub. Clone the repo or copy the file directly into your Claude Code skills directory.
npx versuz@latest install vivekkarmarkar-claude-code-os-skills-checkweather-next-day-vivekgit clone https://github.com/VivekKarmarkar/claude-code-os.gitcp claude-code-os/SKILL.MD ~/.claude/skills/vivekkarmarkar-claude-code-os-skills-checkweather-next-day-vivek/SKILL.md# Check Weather Next Day — Vivek's Tomorrow Weather Briefing Get a full weather briefing for TOMORROW, tailored to Vivek's preferences, walking habits, running style, and Claude Code remote-control workflow. ## Arguments `<location>` — City name or coordinates for where you WILL BE tomorrow. If no arguments, ASK: "Where will you be tomorrow?" Examples: - `/checkweather-next-day-vivek Iowa City` - `/checkweather-next-day-vivek NYC` - `/checkweather-next-day-vivek 41.66,-91.53` - `/checkweather-next-day-vivek` — asks "Where will you be tomorrow?" ## Vivek's Personality & Preferences These are hard constraints. Always apply them: - **NOT a morning person.** Never recommend anything before 11 AM. Morning options are mentioned only to explain why they're skipped. - **HATES hot weather.** Sunny, 70°F+ is miserable. The hotter and sunnier, the worse. - **Loves cool weather.** 40°F with mild sun and no wind is PERFECT. That's the gold standard. - **Cold tolerant.** Can walk and run comfortably in 20°F. Cold + wind is manageable. Cold is never a dealbreaker. - **Wind sensitive for phone use.** 15+ mph wind makes typing on a phone annoying. Factor this into remote-control recommendations. - **Daily step target: 20,000+ steps.** This is non-negotiable. The strategy must achieve this. - **Runs regularly.** Getting back into shape. Uses Strava. Prefers evening runs. ## Workflow ### Step 0a: Health Check Check that the Open-Meteo MCP server is healthy: - Verify `mcp__open-meteo__weather_forecast` is available - If NOT healthy, tell the user to reconnect and STOP - If healthy, proceed ### Step 0b: Resolve Location If the user didn't provide a location, ASK: **"Where will you be tomorrow?"** Wait for the answer. Do NOT assume they will be where they are today. Once provided, map to lat/lon: - Iowa City → 41.6611, -91.5302 - NYC / New York → 40.7128, -74.006 - For other cities, use the `mcp__open-meteo__geocoding` tool to resolve ### Step 1: Fetch Weather Data Call `mcp__open-meteo__weather_forecast` with: - Coordinates from Step 0b - `current`: DO NOT request current conditions (we don't care about right now) - `hourly`: temperature_2m, apparent_temperature, precipitation_probability, rain, snowfall, wind_speed_10m, relative_humidity_2m, weather_code - `temperature_unit`: fahrenheit - `wind_speed_unit`: mph - `timezone`: appropriate for the location (America/Chicago for Iowa City, America/New_York for NYC, etc.) - `forecast_days`: 2 **Filter the hourly data to TOMORROW only.** Discard today's data entirely. ### Step 2: Present Tomorrow's Conditions Summary Show a quick summary of tomorrow's day: 1. **High / Low** in both °F and °C 2. **Wind Range** in mph 3. **Rain Chance** — peak percentage 4. **Snow** — amount or "None" 5. **Overall Conditions** — human-readable summary ### Step 3: Present Hourly Breakdown Show a table from 10 AM to 10 PM (Vivek's active hours) with: - Time, Temp (°F/°C), Feels Like, Rain %, Wind, Humidity, Conditions - **Bold** the best windows for Vivek's preferences ### Step 4: Walking Strategy (20K+ Steps) Analyze the hourly data against Vivek's preferences and produce: - **Rate the day** against Vivek's ideal (40°F, mild sun, no wind) - **Identify the best window** — when conditions closest match his preferences (cool, calm, not sunny) - **Identify the worst window** — when conditions are furthest from his preferences (hot, sunny, windy) - **State "since you're not a morning person"** and explain why morning is skipped - **Give a concrete hour-by-hour strategy** to hit 20K+ steps, concentrated in the best windows - **Suggest what to wear** — windbreaker vs rain jacket vs layers vs t-shirt - **Compare to his ideal** — how close is tomorrow to the 40°F-mild-sun-no-wind paradise? ### Step 5: Running Strategy Analyze conditions for three run types: - **Short run (3-5 miles):** Best time, expected conditions, what to wear - **Long run (8-12 miles):** Start time, how conditions change over the run, route strategy (wind direction) - **Ultra (15+ miles):** Is it viable tomorrow? If yes, start time and hydration notes. If no, explain why (humidity, heat, etc.) For each: - **State "since you're not a morning person"** and confirm evening is better - Consider how temperature CHANGES during the run (descending temps = good for Vivek) - Note humidity's effect on cooling efficiency - Wind direction strategy (into wind first, tailwind home) ### Step 6: Remote Control Claude Code Strategy Analyze conditions for phone-based Claude Code work while walking: - **Phone screen readability** — sun glare, rain on screen, brightness needed - **Typing comfort** — finger temperature, wind resistance on phone - **Battery strategy** — screen brightness recommendations, estimated drain - **Divide the day into phases:** - Desk Mode (when conditions are worst for Vivek's outdoor preferences) - Light Remote Control (read-heavy, wind too strong for typing) - Medium Remote Control (back-and-forth conversations) - POWER MODE (ideal conditions — go hard on complex Claude Code work) - **Route recommendations** for the location — wind-sheltered paths for windy hours, open paths for calm hours - **Match remote control intensity to weather phases** — don't fight bad conditions, wait for good ones ### Step 7: Meta-Insight Close with one sentence connecting tomorrow's weather pattern to how Vivek should structure his time. Something that ties the weather, the walking, the running, and the Claude Code work into a unified daily strategy. ## Rules 1. **Always check MCP health first.** Don't proceed with stale or broken data. 2. **Always ASK where the user will be tomorrow.** Do NOT assume same location as today. 3. **Always apply Vivek's preferences.** Cool > warm. Evening > morning. No heat. No early mornings. 4. **Be opinionated.** Don't hedge with "you could do X or Y." Tell him what to do and when. 5. **Be specific about times.** "Go at 7 PM" not "go in the evening." 6. **Temperature always in both °F and °C.** Never just one. 7. **Compare to his ideal.** Every briefing should reference the 40°F/mild-sun/no-wind benchmark. 8. **Don't sugarcoat bad weather.** If it's 90°F and sunny, say "stay inside, this is your nightmare." 9. **Remote control strategy is mandatory.** This isn't a generic weather app — it's a briefing for someone who works from their phone while walking. 10. **Filter to TOMORROW only.** This skill is about planning ahead, not about right now.