hertz, post-clear. State + one question.

#198: still OPEN at ff4b405d, run 34090992646 all 5 checks SUCCESS (changes, traceability,
lint, unit Linux kitsubito, unit Windows hfenduleam). Nothing changed since my 06:43Z green
report. Waiting on your ff land. No box touched — an open PR is a battery on both.

RIDER 1 census RE-DERIVED just now at ff4b405d by grep alone, not recalled from the notes:
- flattened-whitespace `"daemon", "start"` matches 8 binaries in crates/spt/tests
- extracted the package(spt) HEAVY member list from .config/nextest.toml:308 (44 names)
- set-differenced them: 7 not yet HEAVY -- boundary_events_e2e, broker_stop_endpoint_deny_e2e,
  daemon_stop_convoy_e2e, io_events_poll_e2e, io_events_undriven_kinds_e2e, midturn_span_e2e,
  now_signal_delta_e2e. (job_escape_e2e is the 8th, already HEAVY for a PTY reason.)
Same 8/7 and same names as the pre-clear measurement. Two independent derivations agree.

QUESTION, and the only thing I am blocked on: my notes say both riders ride the IR-37 thin PR
AFTER W2 lands. Do you want RIDER 1 authored NOW as a prepared commit on a branch off ff4b405d
(criterion widened by the pair + all 7 into BOTH <HEAVY> strings + the measurement recorded,
one commit per your 06:46Z constraint), or held until W2 is in so todlando's hand-add of
webserve_attachment_e2e and my 7 go into the strings without a conflict?

I can author it with zero box cost -- the identity assert is a set comparison I can compute by
grep on both sides, as above. What I cannot do without a box is BUILD it, so a prepared commit
would be authored-but-unverified until the lane opens. Your call which risk you prefer.

Idle otherwise. Nothing else of mine is in flight.