Add probe P8: Wait and Total Time Elapsed under Set Slow Motion #122

Problem

Bastion on the committed docs/scenario-examples/valid/bastion.json now runs 91 ticks, then stops not_modeled: action wait under a 90% simulation rate, rule 218 initialize effects, action 2 (Wait(Random Real(1, 2), Ignore Condition)).

Bastion sets Set Slow Motion(90) in rule 187 and never restores it, so every later Wait runs under a 90% rate. #120 correctly refuses to guess how seconds map to ticks (and how Total Time Elapsed advances) under a non-100 rate. No capture measures it. Probes P0–P7 deliberately never call Set Slow Motion.

Scope

Add one probe, P8, that measures the slow-motion regime in the real client, following docs/capture-convention.md and the existing probe layout (probes/p8-slow-motion-wait/: probe.workshop, probe.zh-CN.ow, expect.json, empty capture.json):

• For factors 50 and 90 (and a 100 control in the same run), measure: elapsed Total Time Elapsed for Wait of 0.016, 0.1, 0.5 and 1.0 s; whether wait rounding (P1) is unchanged; whether Total Time Elapsed itself advances at the factor or at wall rate; and the interval between rule starts.
• Print times in ms (_ms fields) per the P0–P4 convention. Reset the factor to 100 at the end.
• Update docs/procedure.zh-CN.md: the probe table and the common-step note that says probe files never call Set Slow Motion (state that P8 is the sole exception and must be the only rule loaded).
• Do not claim a capture. The client run is done by the owner or by an agent following wrightkit-run-workshop-probe; capture.json stays at 0 runs until athena ingest records them.

Acceptance

• athena validate (or the repo's probe validation) passes for P8.
• The probe's end line and expect.json lines are consistent with the printed fields.
• Docs updated; no change to simulator behavior in this issue. A follow-up issue models the measured mapping once captures exist.