CROUCH complete: full cycle + MP replication verified

The WIP's 'pose does not hold' was a chain of bench-instrument errors, not a
code bug: every capture ran in COCKPIT view (the pilot cannot see their own
legs; the eye-height residual masked as reversion).  Joint probes prove the
park holds indefinitely (knee 1.138, root -2.219 steady); the 2-node bench
shows the observer's replicant fully crouched and held (duckmpA_031 -- the
type-3 state record carries it with zero new replication code); the second
scripted press (new BT_BTNTEST2 env) verifies RISE -> standing zeros.
MP button delivery confirmed mode-mask-clean (the one miss was round-start
jitter).  Diags added, all BT_DUCK_LOG-gated: SetLegAnimation re-arm tracer,
1 Hz joint probe, RIO press mode-mask, BT_TREE_LOG topology dump, and the
squat-park log.  RESIDUAL filed: pilot's own cockpit eye does not ride the
root drop (DPLEyeRenderable chain composition; cosmetic).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

context/locomotion.md

@@ -311,6 +311,23 @@ resolved with gait v5, no shadow code changed — though (a) may have contribute
 T4]; (c) genuine slope burial when the quad is flat/mis-tilted (the depth-bias margin is finite).
 Check `[shadowobj]` tag lines and `[sync]`/`BT_SYNC_LOG` before touching bias/tilt.
 
+## CROUCH -- COMPLETE 2026-08-06 [T1 decode / T2 full-cycle + MP benched]
+The manual's CROUCH button (streamed 0x13; glass key F4), end to end: DuckRequest (mech id 0x1a
+@0049fa00) latches duckState@0x398 -> the master posture block (mech4, from the dark master-perf
+@0x4a9cf0-0x4aa0af): posture selector @0x3f8 (mode 9/10/2/3/4 block, NOVICE simLive lockout,
+myomer-alive gate) -> standing: SetLegAnimation(2) ('sqd', squat down; LegClipFinished case-2
+PARKS the pose + leg alarm 1) / ducked: SetLegAnimation(3) ('squ', rise; case-3 alarm 0);
+stability alarm flips; ForceUpdate(8)+(1) ships the type-3 state record -- peers pose the squat
+for FREE (screenshot receipt: the crouched crimson madcat from the observer's cockpit,
+duckmpA_031). Clip data verified from BTL4.RES (root joint drops -2.22, knees +-1.14; squ/sqd
+are exact mirrors; interior 'squi/sqdi' variants ship for all 8 chassis). Full solo cycle
+benched (BT_BTNTEST + BT_BTNTEST2): squat -> hold (joints steady for minutes) -> rise ->
+standing zeros. Myomer speed-cripple feeder (@004b8be3) still deferred [T3] -- see mech4 banner.
+RESIDUAL (open-questions): the pilot's own cockpit EYE height does not ride the root drop
+(the interior clip authors joint-0 translation for exactly that; the DPLEyeRenderable chain
+composition needs a look).  Bench traps for reruns: MP button delivery can miss if pressed
+during round-start jitter (press >= poll 900 after GO); diags all BT_DUCK_LOG-gated.
+
 ## Key Relationships
 - Detail: `docs/P3_LOCOMOTION.md`. Uses: [[asset-formats]] (SKL/ANI), [[decomp-reference]] (offsets).
 - Feeds: [[combat-damage]] (collision→damage), [[rendering]] (shadow/visual-conform).

context/open-questions.md

@@ -1142,3 +1142,10 @@ Gitea #116 and `experience-levels.md`.
 - **des_night-family pages have subtle fog deltas** (both sets near-black); arena/polar night
   pages carry the dramatic 5↔40 near-plane swing. If testers report "searchlight does nothing"
   on some maps, it's page authoring, not the mechanism.
+
+## Crouch residual (2026-08-06 — the feature itself is LIVE, [[locomotion]] §CROUCH)
+- **Pilot eye height does not ride the root drop.** The interior squat clip animates joint 0
+  (ball+translation, −2.22) precisely so the cockpit eye lowers; the joint holds (probe-verified)
+  and the eye's segment chain passes through jointlocal, yet the cockpit view returns to standing
+  height after the squat parks. Suspect: DPLEyeRenderable's parent-chain composition not
+  including the BallTranslateJointRenderable's live matrix. Cosmetic (body + replication correct).