Fetches live observations and historical day/month/year statistics from WeatherFlow Tempest stations or devices using the Tempest REST API. The skill provides machine-friendly JSON output and an optional short human summary, and supports metric/us units and targeted queries by station or device ID.
Use this skill when the user asks for local or backyard weather readings from a Tempest station, wants wind/gust/rain/lightning data from a specific device, or needs historical station statistics for a given day, month, or year. Triggers include queries mentioning "tempest weather", "Tempest station", "backyard weather", or requests for station stats.
scripts/get_tempest_weather.py for live observations and --stats modes for historical rows. (has_scripts=true)references/tempest-api.md mapping observation indices and response notes. (has_references=true)Best suited for agents with HTTP/webfetch and Python runtime support (e.g., Claude Code, Codex/Copilot-style integrations that can run helper scripts or call web APIs).
Tempest Weather fetches current observations and historical stats from a WeatherFlow Tempest station via REST API. The Python script is well-written with proper argparse, unit conversion, and retry logic; it exited cleanly with an error when no token was provided (expected in dry-run). The bash packaging script timed out. No security concerns — tokens are env-based and redacted in output URLs.
Clean, well-structured skill for a specific hardware niche. No security issues. Scripts behave correctly given missing config. The package_clean.sh timeout is likely due to missing package_skill.py in the sandbox.
