What problem does it solve? Translating natural-language weather requests like "heavy rain at sunset" into CARLA's 14 WeatherParameters fields is error-prone guesswork. This Skill maps plain-language requests to exact preset values or field overrides, applies them to a running CARLA server, and reads the result back for verification. ## Core Features & Use Cases - Natural-language to parameter mapping: Converts requests like "light fog at night" or "dust storm" into exact WeatherParameters using an authoritative vocabulary and preset matrix covering all 23 built-in presets. - Preset, tweak, or compose: Apply a named preset directly, start from a preset and override specific fields with --base, or compose all 14 fields from scratch including sun angles, fog, puddles, and scattering scales. - Verified application: Every change is re-read from the server into a VERIFY block, since set_weather returns nothing, so the applied state is confirmed against the request. - Use Case: A user asks to make it "foggy night, calm" in their simulation. The Skill applies set --base ClearNight --fog-density 90 --fog-distance 10 --wind-intensity 5 and verifies sun_altitude is -90 and fog_density is 90. ## Quick Start Ask the agent to set the weather on your running CARLA server, for example "make it rain hard at sunset", and it will apply the matching preset and verify the result.
