[khiet]

Summary

Design only: a mockup comparing three rain versions and ADR 0103 recording grip only. Nothing in the sim or client changes.

Weather pick: Rain (starts on Dry)
  seed picks the lap, from lap 2
    first kart onto that lap -> wetness 0 to 1 over 20 s, held to the flag
      steering x 0.92, acceleration x 0.92, braking x 0.80
      wash-out + 0.05 rad, to at most 0.18
      top speed cap and wear unchanged
      bots keep today's rules

Build issues to file after this merges: the sim and the pick; the look; the playtest and retune. The build waits for the playtest ADR 0085 and ADR 0095 are pending.

Evidence

• Before: no record of whether rain should exist or what it does.
  After: docs/mockups/rain-on-a-seeded-lap.html and docs/adr/0103-rain-changes-grip-on-a-lap-the-seed-picks.md.

A throwaway trial of about 400 eight-bot races, not committed, scaled the grip factors mid-race:

Rain level	Karts past the end cap, 5 laps (of 512)	Ace stops a kart after the rain, 5 laps
Dry control	39	1.26
Mild (chosen)	59	1.56
Heavy	104	not averaged; per cell in the mockup

It is enough to compare levels, not to tune. It did not race the ramp, other seeded laps, Tyres Off packs or wets, and did not count missed pit lane forks; the ADR leaves those to the build's sweep. The pack picture is a hand-drawn sketch, so spray still needs a staged eight-kart race.

pnpm test: 1146 passed, including the repo-checks ADR list. pnpm lint: clean.

Screenshots:

Pack sketch, fully wet with the spray budget

Merge Danger

Door: two-way

Docs only; a later ADR can supersede this one before anything is built.

Blast Radius: none

Adds one slug to the repo-checks ADR list. No sim, client or replay change.