acdream/docs/research/2026-05-23-a6-p3-issue98-comparison-harness-findings.md

# A6.P3 #98 — Comparison harness shipped, root cause identified

**Session:** 2026-05-23 evening (continuation of full-day session)
**Worktree:** `C:\Users\erikn\source\repos\acdream\.claude\worktrees\strange-albattani-3fc83c`
**Branch:** `claude/strange-albattani-3fc83c`

Read this AFTER the morning's handoff doc
([`2026-05-23-a6-p3-issue98-harness-handoff.md`](2026-05-23-a6-p3-issue98-harness-handoff.md)) —
this picks up from "Option A: build the side-by-side comparison harness" and
documents the FIRST evidence-driven step in the saga.

---

## TL;DR

**Updated 2026-05-23 evening v3: NEW root-cause hypothesis identified —
STALE RAMP CONTACT PLANE causes per-tick Z drift, which is what makes
the cottage-floor cap reachable in the first place.**

- Player position at cap: world (141.5, 7.2, 92.7). The cellar ramp's
  actual world XY is X=[129.7, 131.3] — the player is **10 meters away
  from the ramp** in cell-local space.
- Body's contact plane: ramp's plane (n=(0, 0.719, 0.695), d=-69.5035).
  Stale; should be the flat cellar floor (n=(0,0,1)).
- AdjustOffset projects forward motion along that stale ramp plane.
  Mathematically: requested delta (+0.0266, -0.4022, 0) → projected
  (+0.0266, -0.1943, +0.2010). **+0.2010 m of Z lift per tick.**
- After enough horizontal-walking ticks, the head sphere rises to
  Z=94 and hits the cottage floor's downward-facing back-face polygon.
  Cap fires.
- The cap is a SYMPTOM. The root cause is the contact plane not
  refreshing when the player walks off the ramp onto the flat cellar
  floor. Retail must re-find the walkable plane each tick; we're
  keeping the stale ramp seed.

**This explains why six prior fix attempts missed.** Step-up,
AdjustOffset projection, SidesType, edge-slide, +X residual — all
were investigating the cap event mechanics, not the upstream Z drift
that made the cap reachable. The harness convergence (Section "What
shipped 2026-05-23 evening v2") is still valuable as the deterministic
reproduction infrastructure; the new hypothesis is the **next** thing
to verify against that infrastructure.

(Sections below preserve the evening-v2 arc for context: apparatus +
cap-event reproduction.)

- **Evidence-driven apparatus shipped.** `PhysicsResolveCapture` writes one
  JSON Lines record per player ResolveWithTransition call when
  `ACDREAM_CAPTURE_RESOLVE=<path>` is set. 41,228 records from a single
  cellar-walk session.
- **Comparison test reproduces the cap divergence on the first try.** The
  new `LiveCompare_*` tests in `CellarUpTrajectoryReplayTests.cs` load three
  representative records (spawn, on-ramp, first-cap) and replay them
  through the harness engine. Spawn + on-ramp PASS bit-perfect; first-cap
  FAILED with a clear divergence — the right divergence.
- **Root cause identified: the cottage GfxObj was missing from the harness.**
  Live cap attributes the blocking entity to `obj=0xA9B47900` — a
  landblock-baked static building. The cottage's floor polygons live in
  this GfxObj's polygon table (registered as a ShadowEntry), NOT in any
  cottage CELL.
- **Apparatus convergence (v2 update).** With the cottage GfxObj
  `0x01000A2B` extracted via the new `ACDREAM_DUMP_GFXOBJS` infrastructure
  and registered as a ShadowEntry in `BuildEngineWithCellarFixtures`, the
  harness now reproduces the live `cn=(0,0,-1)` cap exactly. The
  full per-field round-trip reveals one residual: live preserves
  +0.0266 m of +X motion through the cap; harness blocks all motion.
  That's the next investigation target — see the "Residual divergence"
  section below.
- **Not a step-up / AdjustOffset bug.** The head sphere (top at Z=foot+1.2)
  hits the cottage floor at Z=94.0 from BELOW. Math: cap at foot Z=92.74
  matches 94.0 − 1.2 = 92.80. Confirmed by user reporting same cap when
  JUMPING in the cellar (purely vertical motion). The retail comparison
  question is now sharpened to "how does live's post-cap edge-slide
  preserve the +X component that the harness drops?"

---

## What ran this session (chronological, 3 commits)

| Commit | What |
|---|---|
| `fb5fba6` | Apparatus: `PhysicsResolveCapture` static class + JSON Lines writer + body snapshot record + capture probe in `ResolveWithTransition` + smoke tests (capture writes when IsPlayer + enabled, skips otherwise) |
| `44614ab` | Comparison test: 3 fixture records sampled from live capture + 3 `LiveCompare_*` tests + diagnostic dump that prints cell polygons in world frame |
| `0f2db62` | Converted FirstCap test to documents-the-bug pattern (passes while harness lacks cottage GfxObj; fails when added) |

Live capture launches:
- `launch-a6-issue98-capture.ps1` — first capture run (no probes beyond cell-transit). Produced `a6-issue98-resolve-capture.jsonl` (12 MB, 5789 records when checked mid-session, finished at 91 MB / 41,228 records).
- `launch-a6-issue98-polydump.ps1` — second capture with `ACDREAM_PROBE_POLY_DUMP`, `ACDREAM_PROBE_PUSH_BACK`, `ACDREAM_PROBE_RESOLVE`, `ACDREAM_PROBE_INDOOR_BSP`, and `ACDREAM_DUMP_CELLS` covering 0xA9B40140-0xA9B4014F. Produced `a6-issue98-resolve-capture-2.jsonl` (135 MB, 70,572 records) plus 16 cell-dump JSON fixtures and a launch log with 214 [poly-dump] entries.

---

## The apparatus (committed code)

### `PhysicsResolveCapture` ([`src/AcDream.Core/Physics/PhysicsResolveCapture.cs`](../../src/AcDream.Core/Physics/PhysicsResolveCapture.cs))

Static module. When `ACDREAM_CAPTURE_RESOLVE=<path>` is set, every player-side
`PhysicsEngine.ResolveWithTransition` call appends one JSON Lines record:

```json
{
  "tick": 0,
  "timestampMs": 40919993,
  "input": { ... full inputs ... },
  "bodyBefore": { ... full PhysicsBody snapshot ... },
  "result": { ... full ResolveResult ... },
  "bodyAfter": { ... full PhysicsBody snapshot ... }
}
```

Filtered to `IsPlayer` mover flag so NPC / remote DR calls don't pollute.
Thread-safe writer with per-record flush. Process-exit hook for clean
shutdown.

### Comparison harness ([`tests/AcDream.Core.Tests/Physics/CellarUpTrajectoryReplayTests.cs`](../../tests/AcDream.Core.Tests/Physics/CellarUpTrajectoryReplayTests.cs))

Three `LiveCompare_*` tests + one diagnostic dump:

| Test | Outcome | Meaning |
|---|---|---|
| `LiveCompare_Tick0_Spawn` | PASSES | Spawn at Z=92.5333; engine matches live bit-perfect |
| `LiveCompare_Tick376_OnRamp` | PASSES | Player on ramp at Z=91.49; ramp walkable polygon hydrates correctly, engine reproduces live |
| `LiveCompare_FirstCap_HarnessMissesCottageFloorBecauseCottageGfxObjNotRegistered` | PASSES (documents the bug) | Live cap at Z=92.74 with cn=(0,0,-1); harness does NOT reproduce because cottage GfxObj isn't registered |
| `LiveCompare_FirstCap_DiagnosticDump` | PASSES (probe-only) | Prints cell polygons in world frame + enables every probe — captured stdout shows harness BSP query path |

The diagnostic dump test runs the cap replay with `[poly-dump]`, `[push-back]`,
`[indoor-bsp]`, `[step-walk]` probes ALL enabled. The captured stdout shows:

```
[cell-dump] 0xA9B40147 resolved-poly-count=37
  poly id=0x0018 ... worldVerts=[(140.12,11.50,90.95),...(142.10,11.50,90.95)]   ← cellar floor
  poly id=0x0001 ... worldVerts=[(142.10,11.50,93.80),...(140.50,8.70,93.80)]   ← cellar ceiling

[cell-dump] 0xA9B40143 resolved-poly-count=14
  poly id=0x0004 ... worldVerts=[(136.70,3.90,94.00),(140.50,3.90,94.00),(140.50,8.70,94.00)]   ← cottage floor (triangle)
  ... more cottage floor triangles, all at world Z=94.00 ...

[other-cells] primary=0xA9B40147 iter=0xA9B40143 wpos=(141.605,7.097,93.351) result=OK poly=n/a
[other-cells] primary=0xA9B40147 iter=0xA9B40146 wpos=(141.605,7.097,93.351) result=OK poly=n/a
```

Both other-cells iterations return OK — the cottage floor polys in
0xA9B40143 don't extend to the sphere's XY (X=141.39 > rightmost-vertex
X=140.50). So the harness sees no collision, even though the live engine
does.

---

## How we identified the missing object (it's NOT a cell)

The second capture pass enabled `ACDREAM_PROBE_RESOLVE=1`, which logs
each call's hit details including the entity guid of the blocking object.
The cap event prints:

```
[resolve] ent=0x000F4240 in=(141.605,7.304,92.656) tgt=(141.624,6.875,92.656)
  out=(141.605,7.304,92.656) ok=True groundedIn=True cp=valid
  hit=yes n=(0.00,0.00,-1.00) obj=0xA9B47900 walkable=True
```

**obj=0xA9B47900** is in the landblock-baked static range (0xA9B47XXX
guids belong to landblock 0xA9B4's static objects). This is the cottage
BUILDING as a GfxObj registered as a ShadowEntry on the landblock —
NOT a cottage cell.

The harness's `BuildEngineWithCellarFixtures` loads three CELL fixtures
(0xA9B40143, 0xA9B40146, 0xA9B40147) but **does not register any
landblock-baked static**. There IS a `RegisterStairRampGfxObj` helper
that constructs ONE polygon (the ramp), but it's commented out today.

So the missing apparatus is: register the cottage GfxObj as a ShadowEntry
with its FULL polygon table — ramp + walls + floor + ceiling. Once
registered, the harness's multi-cell BSP iteration's
`FindObjCollisions` will query the GfxObj's BSP and find the cottage
floor polygon's downward-facing plane just like live.

---

## The cap geometry (math)

Live capture analysis confirmed the sphere physics:

- Foot sphere center at world Z = foot_z, radius 0.48m
- Head sphere center at world Z = foot_z + sphereHeight = foot_z + 1.2m
- Head sphere top at Z = foot_z + 1.2 + 0.48 = foot_z + 1.68m

Cap point in live capture: foot_z = 92.7390 (from tick 1183).
Predicted head sphere position: head center = 93.9390, head top = 94.4190.

The cottage floor is at world Z = 94.0 (from cell 0xA9B40143's poly 0x04
worldVerts: `(136.70,3.90,94.00)`, etc.).

**Head sphere center at Z=93.94 is BELOW the cottage floor at Z=94.0 by 0.06.**
**Head sphere top at Z=94.42 is ABOVE the cottage floor by 0.42.**

The head sphere PENETRATES the cottage floor. BSP push-back direction
is the negative of the polygon's outward normal (which is +Z facing UP),
so push-back direction is −Z (pushes sphere DOWN). That matches the
live cn=(0,0,-1).

The "exact" cap position: foot_z when head center is at Z=94.0 (just
touching). foot_z = 94.0 − 1.2 = 92.80. The observed cap at foot_z=92.74
is ~0.06 below the predicted (push-back includes epsilon and walk-interp
adjustments).

---

## User's confirming observation

> "I noticed a thing. When I jump in the cellar, I'm getting blocked at
> the same height (I think) as I am when running up the stairs."

This is the key observation that nailed the diagnosis. **Jumping is
pure vertical motion** — no ramp slope, no AdjustOffset projection. If
the cap fires on a pure jump, the obstruction must be a horizontal
geometric obstacle at the cap height. That immediately rules out every
step-up / AdjustOffset hypothesis from the prior 6+6 saga and pinpoints
the bug as a head-sphere head-on collision with a cottage-floor
polygon facing DOWN.

---

## What's NOT yet known

1. **Why retail doesn't have this cap.** Either:
   - (a) Retail's cottage GfxObj has a HOLE in the floor above the ramp
     (cottage floor polygons stop at the ramp opening; our dat-read
     produces a contiguous floor)
   - (b) Retail's BSP query treats single-sided polygons correctly
     (cottage floor's SidesType allows collision from +Z side only,
     not from −Z side; we treat it as both-sided)
   - (c) Retail uses portal-aware collision: when the sphere is inside
     the cellar EnvCell, queries skip polygons that belong to the
     cottage portal's "other side"

   Need a retail cdb trace at the ramp-top to disambiguate.

2. **The cottage GfxObj's full polygon list.** We have the ramp polygon
   (poly 0x0008 in the cottage GfxObj, normal (0,-0.719,0.695)) and we
   know the floor polygon is at Z=94.0 with normal (0,0,-1) or (0,0,+1).
   We do NOT have:
   - the full polygon list of GfxObj 0xA9B47900
   - the cottage GfxObj's id, BSP root, or scale/rotation

   These can all be extracted by enabling `ACDREAM_PROBE_BUILDING=1` for
   a future capture — the `[resolve-bldg]` probe dumps per-poly geometry
   when a building shadow entry is hit.

3. **`ACDREAM_PROBE_POLY_DUMP` doesn't fire for the cottage hit.** The
   [poly-dump] probe is wired into `AdjustSphereToPlane`, but the
   cottage-floor collision goes through `FindObjCollisions` →
   `BSPQuery.FindCollisions` on the GfxObj's internal BSP — a different
   code path. Future probing should use `ACDREAM_PROBE_BUILDING` instead
   to capture the per-object collision details.

---

## Next-session pickup

### What shipped 2026-05-23 evening v2 (post-prior-section)

Three commits land apparatus convergence on the cap event:

| Commit | What |
|---|---|
| `cc3afbc` | **GfxObj dump infrastructure.** Mirrors `ACDREAM_DUMP_CELLS`: new env var `ACDREAM_DUMP_GFXOBJS` triggers `PhysicsDataCache.CacheGfxObj` to write the full resolved polygon table as JSON, suffix `.gfxobj.json` so dumps don't collide with cell dumps in the same dir. New `GfxObjDump` DTO + `GfxObjDumpSerializer` parallel to `CellDump`; round-trip tests cover Capture / Write / Read / Hydrate; the Hydrate path constructs a synthetic single-leaf BSP for query coverage. |
| `97fec19` | **Harness reproduces the cottage-floor cap event.** `BuildEngineWithCellarFixtures` now registers a stub landblock 0xA9B40000 (TerrainSurface at z=-1000) so `TryGetLandblockContext` succeeds at the cellar XY, plus a new `RegisterCottageGfxObj` helper that loads the dumped cottage GfxObj fixture, hydrates it with synthetic BSP, and registers as a ShadowEntry at world (130.5, 11.5, 94.0) with 180° Z rotation — matching production's `GameWindow.cs:5893` registration shape for landblock-baked statics. The cottage fixture (74 polys, 6 downward-facing floor triangles, BSP radius 13.989 m) lives at `tests/.../Fixtures/issue98/0x01000A2B.gfxobj.json`; capture launch script is `launch-a6-issue98-cottage-gfxobj-dump.ps1`. |

Test outcome at apparatus convergence:

| Test | Outcome | Meaning |
|---|---|---|
| `LiveCompare_Tick0_Spawn` | PASS | Spawn round-trip preserved by the new landblock + cottage state |
| `LiveCompare_Tick376_OnRamp` | PASS | On-ramp round-trip preserved |
| `LiveCompare_FirstCap_HarnessReproducesCottageFloorCapNormal` | PASS (NEW) | Harness reproduces the live cn=(0,0,-1) cap-event normal exactly |
| `LiveCompare_FirstCap_ResidualXMotionDivergence_DocumentsNextInvestigation` | PASS (documents-the-bug) | Captures the ONE remaining post-cap divergence: live preserves +0.0266 m of +X motion through the cap (edge-slide along the cottage floor in XY); harness blocks ALL motion. Y and Z agree. |

### The residual divergence (next investigation target)

After registering the cottage GfxObj:

```
Live:    cn=(0,0,-1), position=(141.3865, 7.2243, 92.7390)   ← +X motion preserved
Harness: cn=(0,0,-1), position=(141.3599, 7.2243, 92.7390)   ← X stuck at input
Input:   currentPos=(141.3599, 7.2243, 92.7390)
         targetPos =(141.3865, 6.8221, 92.7390)
         requestedDelta=(+0.0266, -0.4022, 0)
```

The cap-event collision normal matches bit-perfect. Position diverges
in X only. Working hypothesis: live's response to a `cn=(0,0,-1)`
head-bump treats it as a Z-only constraint and edge-slides the
remaining XY component along the cottage floor; harness's BSP path is
rejecting the entire move vector instead of computing a slid offset.

That hypothesis is the next-session investigation target — work the
slide path in `Transition.transitional_insert` / `AdjustOffset` against
the production cap-event call. The new
`LiveCompare_FirstCap_ResidualXMotionDivergence_DocumentsNextInvestigation`
test PASSES today (asserting the current residual) and FAILS when the
divergence closes — that's the signal to flip it into
`AssertCallMatchesCapture` form.

### Alternative pickup move: retail cdb trace at the cottage ramp-top

If apparatus polish is enough and the user wants to widen the question
to "how does retail differ?", attach cdb to a running retail acclient
(see CLAUDE.md "Retail debugger toolchain"), set breakpoints on
`BSPTREE::find_collisions` and `CGfxObj::shadow_find_obj_collisions`,
walk up the cottage ramp, and log every BSP query against the cottage
GfxObj. Compare which polygons retail finds vs which polygons our
acdream engine finds. Retail's trace is the ultimate oracle for the
"how does retail differ?" question — but the apparatus-side X residual
investigation is the more focused, faster-feedback next step.

### Pre-existing test flakiness (out of scope but documented)

While verifying the cottage helper, the full `dotnet test` serial run
produced 8–19 failures across 1192 tests depending on order — the
suite has static-state leakage between test classes (likely from
`PhysicsResolveCapture.CapturePath`, `PhysicsDiagnostics.Probe*Enabled`,
and similar global mutators). The flakiness is **independent of A6.P3**:
stashing the cottage helper out and rerunning produces the same flaky
range. All 21 issue-#98-relevant tests (12 harness + 4
`GfxObjDumpRoundTripTests` + 1 new `PhysicsDiagnosticsTests` + 4
`CellDumpRoundTripTests`) pass deterministically in isolation.

---

## Apparatus that exists to use

| Tool | Location | Status |
|---|---|---|
| `PhysicsResolveCapture` | `src/AcDream.Core/Physics/` | Production-ready; env-var gated; off by default |
| `LiveCompare_*` tests | `tests/.../CellarUpTrajectoryReplayTests.cs` | 4 tests; 1 documents the bug, 3 are matches |
| `live-capture.jsonl` fixture | `tests/.../Fixtures/issue98/` | 3 representative records (spawn, on-ramp, first-cap) |
| `launch-a6-issue98-capture.ps1` | worktree root | Capture-enabled launch (no diagnostic probes) |
| `launch-a6-issue98-polydump.ps1` | worktree root | Capture + poly-dump + push-back + dump-cells launch |
| 16 cell-dump fixtures | `tests/.../Fixtures/issue98/0xA9B4014X.json` | All cells in 0xA9B4014X range from second capture |
| 41K-record live capture | `a6-issue98-resolve-capture.jsonl` (gitignored size) | First capture — full session of cellar movement |
| 70K-record live capture w/ probes | `a6-issue98-resolve-capture-2.jsonl` | Second capture — included poly-dump events |
| `a6-issue98-polydump-launch.log` | worktree root | 56K+ line log with [resolve], [poly-dump], [other-cells], [indoor-bsp] events |

---

## The stale-contact-plane finding — full evidence (2026-05-23 evening v3)

### How the question led to the answer

User asked: "We know how retail OPENs it from above, how hard can it
be to know how to open it from below?" — the implicit question being
"if walking on the cottage floor from above works fine, why doesn't
walking up from below?"

That reframed the investigation. The cottage floor is the SAME
polygon set whether viewed from above (walking on it) or below
(head-bumping it from the cellar). Retail handles both. If our cap
fires from below, what's different about our state?

Tracing the harness's `LiveCompare_FirstCap_DiagnosticDump` output
revealed:

1. **The contact plane the engine started with**: ramp's plane
   `n=(0, 0.7190, 0.6950), d=-69.5035`. From the live capture's
   `bodyBefore.contactPlane`.

2. **Cellar ramp's actual world position**: vertices computed from
   the cellar cell's fixture put the ramp at world
   X∈[129.7, 131.3], Y∈[10.19, 13.09], Z∈[92.5, 95.5]. The ramp is
   in the +Y corner of the cellar, ~1.6 m wide.

3. **Player position at cap**: world (141.5, 7.22, 92.74). 10+ m
   away from the ramp in X.

4. **The +Z drift math**: `AdjustOffset` projects the requested
   motion onto the plane perpendicular to the contact-plane normal:
   - requested = (+0.0266, -0.4022, 0)
   - dot(requested, ramp normal) = 0·0.0266 + 0.719·(-0.4022) +
     0.695·0 = -0.2892
   - projected = requested - (-0.2892)·rampNormal =
     (+0.0266, -0.1943, +0.2010)
   - **+0.2010 m of Z gain per tick**, applied because the contact
     plane the engine believes the player is on is the slope.

5. **The cap math**: foot Z at cap = 92.74. Head sphere center at
   foot Z + sphereHeight 1.2 = 93.94. Head sphere top at
   foot Z + 1.68 = 94.42. **Cottage floor at world Z=94.00.** Head
   sphere top exceeds cottage floor by 0.42 m → cap fires from
   below.

If the contact plane were the flat cellar floor (n=(0,0,1) at
Z=90.95) instead of the ramp, AdjustOffset's projection would
produce zero Z gain (requested motion has no Z component, projection
onto flat-floor plane preserves XY). No drift, no cap.

### Why this fits the user-facing bug

- "Stuck climbing cellar" — the player walks forward, accumulates Z,
  bumps cottage floor, can't progress. Matches what the user sees.
- "Pure jump in cellar caps at same Z" — jumping doesn't refresh the
  contact plane either. Drift continues. Matches.
- "Six prior fix attempts failed" — all attempted to fix the CAP
  mechanics (step-up, slope projection at the cap, edge-slide). None
  questioned why the contact plane was the ramp at all.

### What still needs verification (next session's task)

1. **Chronological evidence**: walk the live capture from the start of
   the cellar session. When did the player last stand on the actual
   ramp? Does `bodyBefore.contactPlane` persist as the ramp's plane
   across many ticks of horizontal walking? Quantify the cumulative
   Z drift.

2. **The walkable-refresh gap**: where in
   `Transition.FindEnvCollisions` / `SpherePath.SetWalkable` /
   related is the contact plane supposed to be refreshed when the
   sphere is over a different walkable polygon? Retail's
   `CObjCell::find_env_collisions` is the decomp anchor — find the
   path that detects a NEW walkable and overwrites the contact
   plane, and find where our engine skips that.

3. **Retail cdb cross-check** (optional, definitive): attach cdb to a
   running retail acclient, walk to a cottage cellar, log the
   contact plane each tick. If retail's contact plane refreshes
   to (0,0,1) when the player walks off the ramp, hypothesis
   confirmed.

---

## Pickup prompt for next session

```
A6.P3 #98 — apparatus convergence landed, NEW root-cause hypothesis
(stale ramp contact plane) needs verification.

Read FIRST (in order, ~15 min):
  1. docs/research/2026-05-23-a6-p3-issue98-comparison-harness-findings.md
     — start with TL;DR (evening v3 update at top), then the section
     "The stale-contact-plane finding — full evidence" near the bottom.
     Skip the middle sections (evening v1 + v2 arcs) unless context is
     needed.
  2. CLAUDE.md "Current A6 phase" block — look for the "Evening v3
     finding" paragraph.
  3. tests/AcDream.Core.Tests/Physics/CellarUpTrajectoryReplayTests.cs
     — the RegisterCottageGfxObj helper + 2 LiveCompare_FirstCap_*
     tests are what you'll iterate against.

State both altitudes (one sentence each):
  Currently working toward: M1.5 — Indoor world feels right
  Current phase: A6.P3 — apparatus convergence shipped (cap event
  reproduces bit-perfect). New root-cause hypothesis: stale ramp
  contact plane causes per-tick Z drift that makes the cap reachable.
  Needs verification.

What was shipped today (3 commits — DO NOT REDO):
  - cc3afbc: GfxObj dump infrastructure (ACDREAM_DUMP_GFXOBJS)
  - 97fec19: Harness reproduces cottage-floor cap (RegisterCottageGfxObj)
  - 7729bdc + (this commit): findings doc + CLAUDE.md updates

The hypothesis with full math:
  - Body's contact plane = ramp's plane (n=(0,0.719,0.695), d=-69.5035)
  - Player position at cap = world (141.5, 7.22, 92.74)
  - Cellar ramp's actual world XY = X∈[129.7, 131.3] — 10m from player
  - AdjustOffset projects requested move along contact-plane perpendicular
  - Per-tick Z gain ≈ 0.201m from slope projection on STALE ramp plane
  - Accumulates over ticks → head sphere reaches Z=94 → bumps cottage
    floor → cap fires
  - If contact plane refreshed to flat cellar floor (n=(0,0,1)) when
    player walks off ramp, no Z drift, no cap

Concrete next moves (in order):

(1) **Verify the hypothesis chronologically.** Walk
    a6-issue98-resolve-capture-2.jsonl (or the cottage capture
    fixture's full file) from the start. Find when the player last
    stood on the actual ramp (within world X∈[129.7, 131.3], Y∈[10.19,
    13.09]). Quantify: how many ticks does the body's contact plane
    persist as the ramp's plane while the player walks horizontally
    away? Compute the cumulative Z drift. Should match observed Z=92.74
    at cap if the hypothesis holds. (Probably 30 min PowerShell jq.)

(2) **Locate the walkable-refresh code path.** In
    src/AcDream.Core/Physics/TransitionTypes.cs, search for where
    Transition.FindEnvCollisions or SpherePath.SetWalkable is supposed
    to detect a new walkable polygon under the sphere and overwrite
    the contact plane. The fix likely lives at the call site that
    EITHER fails to fire OR fires but doesn't replace the existing
    contact plane.

(3) **Cross-ref retail decomp.** acclient_2013_pseudo_c.txt's
    CObjCell::find_env_collisions + the walkable-detection chain.
    Find the path where retail unconditionally replaces
    contact_plane when a new walkable is found. Quote the line
    numbers in the fix commit.

(4) **Implement the fix + verify against harness.** The harness's
    LiveCompare_FirstCap_HarnessReproducesCottageFloorCapNormal test
    currently PASSES asserting the cap reproduces. After the fix,
    if the contact plane refreshes correctly, the cap should NOT fire
    (no Z drift to make it reachable). The test should start FAILING
    — that's the signal the fix works.

(5) **Visual verification (user-side).** Launch acdream live, walk
    into a Holtburg cottage, down to the cellar, then back up. The
    user-facing bug should resolve if the hypothesis is correct.

Decomp grep targets:
  - CObjCell::find_env_collisions
  - CPhysicsObj::find_object_collisions
  - CTransition::find_walkable
  - CSpherePath::set_walkable / walkable_hits_sphere
  - OBJECTINFO::object → contact_plane writes

CLAUDE.md rules apply throughout:
  - NO speculative fixes — the saga's converted to evidence-driven.
    Verify hypothesis with chronological capture BEFORE coding.
  - Visual verification belongs to the user.
  - If the chronological verification (step 1) shows the contact
    plane is NOT actually stale across many ticks, the hypothesis is
    wrong — pivot to retail cdb trace (definitive oracle).

Out-of-scope but observed: pre-existing test suite has 8–19 failures
across runs of the same code due to static-state leakage between test
classes (PhysicsResolveCapture, PhysicsDiagnostics statics). Targeted
issue-#98 tests pass deterministically in isolation. Don't touch the
flakiness this session; it's a separate investigation.

Test baseline: harness's 12 CellarUpTrajectoryReplayTests + 4
GfxObjDumpRoundTripTests + 1 new PhysicsDiagnosticsTests + 4
CellDumpRoundTripTests all pass in isolation. Maintain.

Test baseline: 1178 + 8 pre-existing failures (serial run).
Maintain throughout. The previously-failing
LiveCompare_FirstCap_HarnessMissesCottageFloorBecauseCottageGfxObjNotRegistered
test is now in documents-the-bug form (PASSES while bug exists; FAILS
when fix lands) — flip it when the cottage GfxObj is registered.
```