diff --git a/AGENTS.md b/AGENTS.md
index 591e454..8ac23f7 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -41,3 +41,10 @@ documented in docs/checkpoint/review.md. Pause feature expansion; use the playab
docs/checkpoint/README.md, coverage.md and implementation-plan.md before gameplay
changes. Only evidence probes or directly evidenced corrections belong in this
checkpoint; distinguish static recovery, port tests and original runtime evidence.
+
+Local REA/Ghidra is available through scripts/rea.sh; see docs/rea.md for the
+ignored launcher environment and JDK paths. Existing client MCP registration was
+not changed. Use generated copies, preserve complete Evidence/snapshots, and keep
+static decompilation distinct from source execution. Current C3 continuation:
+docs/checkpoint/control-phase-implementation.md; remaining body/wrapper/camera gates
+are in its final section and the implementation ledger.
diff --git a/README.md b/README.md
index 3dd086a..3232f31 100644
--- a/README.md
+++ b/README.md
@@ -175,3 +175,8 @@ use Wine, DirectX or OpenGL. See `docs/reference-testing.md` for actual results.
All generated extraction output, local SDK/native dependencies, captures and
packages are ignored by the nested local Git repository. Never configure a remote,
upload assets, or overwrite the original installation.
+
+Native reverse engineering uses the local REA/Ghidra launcher separately from
+runtime builds; see [local analysis setup](docs/rea.md). Current source-backed
+cutscene control/body phase findings are in
+[control-phase implementation](docs/checkpoint/control-phase-implementation.md).
diff --git a/docs/architecture.md b/docs/architecture.md
index f35be5b..b59574b 100644
--- a/docs/architecture.md
+++ b/docs/architecture.md
@@ -190,3 +190,16 @@ all physics/AI/render fields. The detailed contract, phase limits and original
component probe are in checkpoint/retained-player-body.md; companion/shadow
resets and full availability tails have separate reports. Uncontrolled physics
and exact appended-helper traversal remain next scheduler boundaries.
+
+## Player control versus frame ownership
+
+PlayerSceneBinding connects authored preparation/transform events to the player
+adapter. Body integration and owned fire run without player control; only input
+and movement math are gated. CAM/MCA/AI preparation has distinct order and operates
+on the prior raw owner, while Force Control affects effective input. Ordinary
+frame/modal/debug pauses stop all these phases. Scripted AI roots/yaw flow into
+physics, and physical roots flow back to director/camera targets. Destination yaw
+and queued work are not cleared by a transform setter. The original full wrapper,
+registry, body response and morph clock are still not reconstructed by this bridge.
+See checkpoint/control-phase-implementation.md. REA tooling is separate from
+runtime architecture; setup/provenance are in rea.md.
diff --git a/docs/checkpoint/README.md b/docs/checkpoint/README.md
index 772aadf..ea5e150 100644
--- a/docs/checkpoint/README.md
+++ b/docs/checkpoint/README.md
@@ -23,6 +23,11 @@ The retained-field continuation adds [property reader](retained-streamer.md),
[player body/scalar boundary](retained-player-body.md) and
[companion/shadow resets](retained-companion.md). It corrects the earlier
availability-prefix interpretation and implements bounded route scalar carry.
+The next C3 continuation is [control phases](control-phase-implementation.md),
+backed by [uncontrolled player](uncontrolled-player.md),
+[physics eligibility](uncontrolled-physics.md) and
+[cutscene preparation](cutscene-preparation.md). [REA setup](../rea.md) records
+local Ghidra/JDK and Evidence/snapshot use.
Evidence labels used throughout:
diff --git a/docs/checkpoint/control-phase-implementation.md b/docs/checkpoint/control-phase-implementation.md
new file mode 100644
index 0000000..6ebaf65
--- /dev/null
+++ b/docs/checkpoint/control-phase-implementation.md
@@ -0,0 +1,80 @@
+# C3 continuation: uncontrolled body and fire
+
+Implementation2026-10-26. Source contracts: uncontrolled-player.md,
+uncontrolled-physics.md and cutscene-preparation.md. This replaces the port's
+whole-player freeze on control loss with a bounded phase split.
+
+LabSimulation.Step consumes prior body state first. Input/controller math runs
+only when Jimmy has effective control; fire ticks independently afterward.
+The existing fresh/retained input ordering remains. Ordinary grounded radius60
+is reasserted before the movement gate, matching the supported wrapper path.
+Grounded fire can expire while mode/jump/hold remain untouched by the skipped
+controller. Queued impulses survive camera preparation and are consumed once by
+the following body pass. Neither control loss nor restoration invents a key flush.
+
+PlayerSceneBinding connects raw-owner preparation events to speed/velocity and
+supported gadget cleanup, and publishes the simulated root while uncontrolled.
+Authored AI transforms update the body too, preventing a later physics publish
+from overwriting the script's teleport. Destination yaw is preserved. Pitch/roll
+remain director-owned; Jimmy's render/controller still uses the existing yaw
+adapter. Initial MCA preparation occurs once, with its first-name guard; later
+shots do not repeat it. AI and CAM use their own callback order. STOP notification
+follows physical preparation. DEFAULT grants control without preparation.
+
+Force Control is a modern effective-input override; raw ownership still selects
+source preparations/music. Noclip keeps its separate flight path. F1/modal/inspect
+boundaries stop the whole frame, so body/fire do not accrue paused time.
+Contact-triggered interactions retain the existing control gate pending C4.
+
+Counters now distinguish total Steps, PhysicsSteps, ControllerSteps and EffectSteps.
+Total Steps also counts modern flight, which bypasses the physical/effect phases.
+The console report exposes these counts; a cutscene's nonzero total count does
+not mean Jimmy accepted input. `--smoke-carl-cutscene` asserts zero controlled
+updates/jumps/turn/speed while physical/effect callbacks continue. The new
+`--smoke-control-phases` seeds an airborne retained-fire fixture, holds movement
+and jump through15 locked frames, checks falling/fire/energy without input, then
+skips/restores control and verifies held-input movement resumes. This is a port
+fixture, not an original end-to-end recording. Headless story pre-roll uses the
+same binding/phase policy.
+
+## Verification and limits
+
+New managed tests cover queued impulse across preparation, uncontrolled fire and
+tutorial immunity, independent grounded resets, CAM/MCA/AI order, body-before-STOP
+observation, raw/effective ownership, later-shot behavior, held jump restoration
+and authored teleport synchronization. Python adds9 source/projection tests.
+The input-only lab replay now uses the same binding and waits through live
+cutscenes with body/fire ticking. Its old coasting turn missed the narrow exit:
+port snapshot root(318.11124,75.02122,770.8816), yaw47.52308, controlJIM1,
+no shot/block and2290 wall hits. Braking with the actual Down input before large
+heading changes restores the route (score150, owned jetpack, one practice No,
+zero jumps,6 wall hits). No physics or task gate was changed to make that replay
+pass. The provisional corner response remains a useful C4 investigation case.
+Native Linux GTX1070 at1280x720 passes Carl/control-phase and F1/F3 tools/menu
+checks with zero enabled Vulkan validation errors. Full integration/package
+results are recorded in progress.md; native Windows remains separate.
+
+REA6.8.0/Ghidra12.1.4 source Evidence is retained under
+artifacts/uncontrolled-phases/rea/. IDs:
+
+| Procedure | Evidence ID |
+|---|---|
+|475CA0 gameplay gate|ev_aa0b7db46894c10d2f28f76d0da1e568ef06253f3d6c8cf6b06fbef299602bab|
+|479BC0 body traversal|ev_56b7e074ec8509d1c5b3576d02361ff051963b777d7ccb8259ed9c65dc2db52f|
+|437940 movement gate|ev_de7097867d129a4232318e667e9496c6868cdb1da4b37a1f3304c94b0b101677|
+|40E050 animated base|ev_03a109302788b85597e253293bbcb535c4080732e41cc828bb33d01d2b2da09a|
+|462520 base body/rotation bridge|ev_f0a2e1fd465472295b272f2fdbfb6be4651aeb8c35d90dc2ca02a495ed359edd|
+
+These are static decompiler observations with unknown C++ types, not original
+source or runtime.421FB0 maps correctly from file21FB0 but lacks an analyzed
+procedure in this Ghidra profile; its pseudocode is unavailable, not evidence
+that the source callback is absent. Existing guarded bytes/vtable/body hash remain
+the authority for that callback. Partial batches preserve successful Evidence.
+
+Remaining: full424600 wrapper timers/UI/sound cleanup, base40E050/462520
+resource/transform/morph policies, collision-triggered ceiling cancellation,
+scooter radius160 and other gadget/class cases, original body registration and
+response, complete registry and engine cadence, shared viewport handoff and
+original whole-game comparison. Calling the entire controller with empty input
+would not implement these missing phases. Existing capsule/yaw/camera adapters
+remain explicit; this chunk does not certify a complete scene scheduler.
diff --git a/docs/checkpoint/coverage.md b/docs/checkpoint/coverage.md
index 9927d9d..8b9aa5b 100644
--- a/docs/checkpoint/coverage.md
+++ b/docs/checkpoint/coverage.md
@@ -12,18 +12,18 @@ runtime comparison. Coverage does not mean implementation completeness.
| ID / system | Recovered contract / anchors | Port implementation | Evidence / remaining gap | Next bounded action |
|---|---|---|---|---|
-| LOOP | Viewport47E4F0 physics before registry473740, including input44DF80; second sync473770; construction order and selective retained-character/new-game ordering. Separate engine morph clock. | Player now integrates prior body state, then fresh input/player/fire or retained player/fire/input. Other systems remain grouped; fixed60Hz policy. | S/M/R. Controlled player partial order implemented; full registry/cadence and uncontrolled-body preparation open. | C3 continue other callbacks and morph/control phases; C4 support/response remains separate. |
+| LOOP | Viewport47E4F0 physics before registry473740, including input44DF80; second sync473770; construction order and selective retained-character/new-game ordering. Separate engine morph clock. | Player now integrates prior body state, then fresh input/player/fire or retained player/fire/input. Uncontrolled body/fire now continue while movement/input are gated. Other systems remain grouped; fixed60Hz policy. | S/M/R. Controlled player partial order implemented; bounded raw-owner preparation implemented; full wrapper/registry/cadence open. | C3 continue other callbacks and morph/control phases; C4 support/response remains separate. |
| INPUT | Game interface+8C slot3C4→44DF80→475A70 reads controlled5099E4 and calls3CC; Jimmy interface+C0→426030. | SDL arrows/jump aliases; control gate, modern debug override; mouse/controller incomplete. | S/M/R, user Windows route. Input occurs in object traversal after physics; device sampling/focus and latency open. | Trace event pump, key/mouse and focus gates alongside retained registry order. |
| MOVE |437940/439900 ground/air arithmetic; physics479D20 consumes queued impulse once before acceleration/gravity. | Controlled player consumes prior body state before callbacks; fresh/retained input order explicit. Fixed dt and custom yaw/model conventions remain. | S/P/M/R. Controlled partial phase split verified; uncontrolled/full registry and body response open. | Continue C3 uncontrolled phases, then C4 support/response. |
| JET |437F90 controller;421FB0 effect; fresh constructor order places fire after Jimmy, both after global physics. States0..5, hold>.4,8s, energy. | Explicit state table with post-collision effect ticking and VR immunity. | S/M/R. Fresh same-pass countdown recovered; retained child lifecycle, sound and expiry→animation open. | Trace retained Jimmy/fire resets and phase end events before scheduler integration. |
| COLL |479D20/47BD80 body response; support and quick magnitude; dynamic boxes; NC-shape registration exclusion. | Source support and box pieces around modern swept24-unit capsule, feet/root grounding and slide. | S/P/M/R. Response/masks/stairs/contact order not fully recovered; source radius60 is not port capsule24. | C4: isolate response and root convention; replace adapter behind route tests. |
| ANIM | Engine morph10002DB0; linear interpolation10002F9A; game logical→OMT sync4624B0. | ASE sample timing, unconditional loop closure, pose/action aliases,1/60 advance. | S/P/M/R. Playback flags, rate, completion callbacks, transition order open. | Recover one-shot/loop/end notifications before more aliases. |
-| SCRIPT | Source AI/CAM/MCA dispatch and chains; task45F990 objective406F50 BEFORE records BEFORE per-entity424. | Bounded synchronous director; baseline objective after records/actor batch. | S/M/R. Script callbacks recovered for route; generic class coverage and reentrancy incomplete. | C1: ordered objective notification; C2: per-actor Carl override. |
-| CONTROL | Global5099E4 null during valid CAM/MCA; PlayerControlled completion target; DEFAULT447D..447F restores camera/UI/music. | Director raw ownership plus explicit debug effective override; player physics stops when uncontrolled. | S/M/R. Null control is not proof all objects/physics stop; scheduler gap remains. | Map raw ownership transitions against phase gates and deferred LOAD. |
+| SCRIPT | Source AI/CAM/MCA dispatch and chains; task45F990 objective406F50 BEFORE records BEFORE per-entity424. | Pre-assignment objective selection and per-actor Carl override implemented; bounded synchronous director. | S/M/R. Script callbacks recovered for route; generic class coverage and reentrancy incomplete. | Continue other class-specific callbacks and reentrancy under the recovered owner/phase boundaries. |
+| CONTROL | Global5099E4 null during valid CAM/MCA; PlayerControlled completion target; DEFAULT447D..447F restores camera/UI/music. | Director raw ownership plus explicit debug effective override; eligible player body/fire continue; only input/controller tail is gated. | S/M/R. Raw/effective control and ordered preparation are implemented; full base/class wrappers and scheduler remain incomplete. | Continue full wrapper/base and collision contact phases around deferred LOAD. |
| CAMERA | Shared viewport; strict>1000 DEFAULT snap; source eye correction/gain; first wall-grid hit then general-grid FIFO; per-level far planes. | Static source query now integrated for the three route scenes; other scenes use provisional sweep. Separate chase/script eyes and merged clip maximum remain. | S/M/R. Query/kernel/counts tested; x87 boundaries, shared handoff/previous-eye and input branches open. | C5 next shared viewport/DEFAULT and complete follow state. |
| AUDIO | CAM416776..E1 stops music; DEFAULT447F12..45 restarts; bank696008 / gate4F8434; handle47D890 polls inherited software duration and repeat+30. | Separate SDL streams, floating-file-duration cutscene completion, independent per-file SFX. | S/M/R (dummy streams). Source duration/allocation/play-failure/bank/attenuation lifecycle incomplete; speakers/native Windows timing not verified. | C6: explicit bank owner and software channel-clock tests; SDL queue-empty is not the source oracle. |
| UI |406F50 selector/timer;4043D9 controlled clock;407070/O edge;404D97 practice resume.44 objective flags distinct from selector. | Timed objective subset, modal practice, modern UiViewport/debug. | S/M/R. Other flags/gadgets/highlights/cursor/energy/layout transitions open; source widescreen not recovered. | Keep modern scale policy separate; recover UI dispatcher/mode gates before expanding. |
-| LOAD | CEditor deferred scene request; selective teardown/reuse; Init stored+11→live+139 resolves old-game deletion; Jimmy/Goddard retained. RETURN root/flags and per-object28→40C. | Reconstructed scene with retained input-order identity and bounded mode/hold/fire container; app-owned presentation. | S/P/M/R. Property stream and bounded scalar/companion resets closed; full body/helper timing and fades remain.50988A is editor R input. | Continue body registration/helper traversal and uncontrolled scheduling; other gadget/level branches remain gated. |
+| LOAD | CEditor deferred scene request; selective teardown/reuse; Init stored+11→live+139 resolves old-game deletion; Jimmy/Goddard retained. RETURN root/flags and per-object28→40C. | Reconstructed scene with retained input-order identity and bounded mode/hold/fire container; app-owned presentation. | S/P/M/R. Property stream and bounded scalar/companion resets closed; full body/helper timing and fades remain.50988A is editor R input. | Continue full wrapper/body registration/helper traversal; other gadget/level branches remain gated. |
| INVENTORY | PostLoad common prefix enables eight owned gadgets4→1, then VR01..08 disables1→4; full tail re-enables the practice gadget. LV1B disables1/7; other wire values preserved. | Availability pass now applied at level entry and verified across TSK tutorial/return; snapshots preserve raw values. | S/M/R. Full selected-gadget/UI and other level tails remain incomplete; bounded route mode/hold/fire carry is implemented. | Keep availability separate from ownership, task changes and arbitrary state carry. |
| SAVE | TSK45FEE0 writer /4603F0 reader; records, arrays, return strings, stored departure root; queued460E70→460EA0. | Original-format round-trip and CLI save, scene snapshots; no full source save UI/load lifecycle. | S/M. Source reader duplicates matching task callback then all-entity420 post-task hook; this is distinct from Jimmy placement PostLoad423A90. Current fresh-director path differs. | C7 after C1/C2; test repeated callbacks before adopting loader behavior. |
| CHECKPOINT | Pickup42AF00 stores+1E38 root/+1E48 level separately from+9A8 RETURN root. | Capture exposed, not used for return; source capture consumers unknown. | S/M. Do not interpret as death/respawn until read consumer found. | Track unresolved consumer; no speculative gameplay. |
diff --git a/docs/checkpoint/cutscene-preparation.md b/docs/checkpoint/cutscene-preparation.md
new file mode 100644
index 0000000..3831750
--- /dev/null
+++ b/docs/checkpoint/cutscene-preparation.md
@@ -0,0 +1,40 @@
+# Cutscene preparation and control handoff
+
+2026-10-26, static EXE evidence. SHA256
+`f5281076e1a2492f2d61f87aae3f4544b6d0f535ebe33f9fbb6f78710fe1aea8`.
+VAs below use400000 preferred base; raw offset=VA−400000 at these sites.
+cutscene_preparation.py records exact bytes/virtual targets; its projection is
+interpretation, not source execution. Primary completed the investigator's helper,
+corrected its sign-extended byte−1 comparison anchor and added tests/report.
+
+| Boundary | Proven order and source sites |
+|---|---|
+|CAM with PlayerControlled=NULL|Guard prior raw owner416635;43841665D→41041666B→flatten logical/cached Euler(0,yaw,0)41667E..4166E5→game118(NULL)4166F4. Target lookup occurs later416754. Preparation can precede a missing target/sound failure.|
+|Initial MCA activation|First target name `none` guard431300 exits before preparation. A non-none name reaches initial target/animation prelude; missing lookup can continue. Prior owner guard431502;410431512→flatten Euler431525..431578→438431598→game118(NULL)4315A7. Per-shot helper invoked431600.|
+|Later MCA shot|CompleteE0→431750; target/sound/animation work, no410/438 or raw control clear. The sound−1 check431771 sign-extends one immediate byte, not a four-byte immediate.|
+|AI PlayerControlled=NULL|Prior raw owner guard40C4AD→41040C4B9→game118(NULL)40C527; no438 or flatten. Named AI ownership remains an older port interpretation requiring broader proof.|
+|DEFAULT|447D5C checks whether Jimmy already owns raw control;447E27 restores via118 without410/438. No new key-release boundary is implemented. Music uses authored ownership independently of debug force.|
+
+Jimmy410→42AFC0 calls43A750 first: clear speed608 and velocity XYZ2C4; stop
+saved sound148; then request complete150(`STOP`) at42AFDE/E5. Body changes precede
+the STOP animation notification. Mode, jump, hold, pending impulse, acceleration
+and fire timer are separate. Existing isolated43A750 probe validates its limited
+write set with invalid sound handle; it does not execute full camera activation.
+
+Jimmy438→42AE00 invokes selected-gadget cleanup178→427FF0 and sets interface1D65
+and global4F8182. Supported no-selection/jetpack cleanup clears selection; other
+gadget cases are not blanket resets (retained-return-cleanup.md). The port exposes
+these ordered boundaries, implements supported player motion/selection and logical
+orientation changes, and leaves other class/UI/resource side effects explicit.
+
+Authored AI movement has separate setters: marker root310 at40C732→target314
+at40C757; yaw40C88F..40C961 uses174→4716D0, writes authored/live Euler and cached
+Euler, and does not write movement destination yaw61C. PlayerSceneBinding applies
+root/yaw to the simulation without erasing pending motion or destination yaw.
+Only actual transform writes emit that event; animation-only changes do not.
+
+The port's skip fast-forward, malformed/reentrant activations, MCA's initial
+animation prelude, all other class virtual effects and full shared camera state
+are not certified by this chunk. Existing named control behavior is preserved,
+not newly recovered here. Three byte/projection tests and coupled managed
+director/body tests guard the implemented ordering and boundaries.
diff --git a/docs/checkpoint/implementation-plan.md b/docs/checkpoint/implementation-plan.md
index 729b8ad..bd17e91 100644
--- a/docs/checkpoint/implementation-plan.md
+++ b/docs/checkpoint/implementation-plan.md
@@ -105,3 +105,20 @@ allowed where it resolves a core-system question, without waiting for every leve
Linux launch from /tmp verifies nondefault tutorial/RETURN scalar boundaries and
body resets. Release validation unavailable; source GPU checks enabled. Native
Windows runtime remains unverified for this chunk. Preserved baseline untouched.
+
+- C3 control continuation: gated movement/input independently of eligible body and
+ fire; explicit CAM/MCA/AI raw-owner preparations and body-before-STOP order;
+ initial MCA target-none returns, later shots do not repeat preparation. Scripted
+ player transforms synchronize into body state. Grounded ordinary radius is outside
+ the control gate. Full wrapper/base/ceiling/class effects remain open.
+- Same binding is exercised by the input-only lab route; modern replay now brakes
+ before large turns instead of coasting past the narrow door. Task/score/inventory/
+ contact assertions remain. The observed corner behavior is retained as C4 input.
+- Local REA/Ghidra/full-JDK setup and batch Evidence/snapshot workflow verified.
+ Five selected source procedures cross-check gates/base calls. Missing analyzed
+ fire entry is recorded; no target execution or client registration change.
+
+- Final C3 integration: full scripts/test.sh passes with93 Python checks,24 movement
+ groups, remaining managed checks and the binding-based input-only lab route,
+ 27 zero-error enabled-validation GPU renders, SDL dummy lifecycle and539 hashes.
+ Native GTX1070 Carl/control/F1/F3 commands pass; native Windows remains separate.
diff --git a/docs/checkpoint/player-phase-implementation.md b/docs/checkpoint/player-phase-implementation.md
index bb2189c..8b5b4c8 100644
--- a/docs/checkpoint/player-phase-implementation.md
+++ b/docs/checkpoint/player-phase-implementation.md
@@ -6,6 +6,11 @@ It does not replace the entire scene scheduler, collision response or animation
clock. Evidence is in `input-frame-bridge.md`, `lab-registry.md`, the new
`player_phase_evidence.py`, and the follow-up `level-game-lifetime.md`.
+Continuation: `control-phase-implementation.md` now separates uncontrolled body/
+fire from controlled movement, with source-specific preparation boundaries. The
+older unresolved-control wording below describes this chunk's original boundary;
+full wrapper/base/morph work and the scene scheduler still remain open.
+
## Source contract and implementation
EXE47E554 calls479BC0 physics before473740 game-object callbacks47E582;473770
diff --git a/docs/checkpoint/retained-player-body.md b/docs/checkpoint/retained-player-body.md
index 006ecd1..8233926 100644
--- a/docs/checkpoint/retained-player-body.md
+++ b/docs/checkpoint/retained-player-body.md
@@ -6,6 +6,10 @@ Preferred-base EXE VAs below map to raw offset VA−400000; the helper independe
maps PE sections. Originals remain untouched. Static recovery, isolated component
execution and port tests are separate evidence categories.
+Continuation: uncontrolled body/fire and cutscene preparation are now implemented
+in a bounded C3 chunk; see control-phase-implementation.md. Full wrapper/base
+policies, body response and scene scheduling remain separate gates.
+
## Closed callback chain
Read `retained-streamer.md`, `retained-return-cleanup.md` and
diff --git a/docs/checkpoint/review.md b/docs/checkpoint/review.md
index 5c7bdbd..c63ef87 100644
--- a/docs/checkpoint/review.md
+++ b/docs/checkpoint/review.md
@@ -10,6 +10,11 @@ availability conclusions are superseded. See `player-phase-implementation.md`,
`level-game-lifetime.md` and the current ledger. The exit review below preserves
the evidence/verification boundary at the original checkpoint milestone.
+The control-phase continuation now separates uncontrolled body/fire from controlled
+movement and implements bounded ordered preparations. Local REA/Ghidra static
+Evidence cross-checks the original gates. See control-phase-implementation.md and
+rea.md; full wrapper/collision/camera work is still outstanding.
+
The initial architecture checkpoint is complete: all requested core systems have
source anchors, state owners/lifetimes, route interactions, baseline comparison,
explicit unknowns and bounded implementation actions. Feature expansion remains
diff --git a/docs/checkpoint/uncontrolled-physics.md b/docs/checkpoint/uncontrolled-physics.md
new file mode 100644
index 0000000..a59bfa3
--- /dev/null
+++ b/docs/checkpoint/uncontrolled-physics.md
@@ -0,0 +1,49 @@
+# Physics eligibility while control is absent
+
+Static C3 follow-up,2026-10-26. Original EXE SHA256
+`f5281076e1a2492f2d61f87aae3f4544b6d0f535ebe33f9fbb6f78710fe1aea8`.
+Preferred-base VAs map to raw offset VA−400000 here; the helper independently
+PE-maps and guards19 instruction sites and6 virtual targets. Investigator supplied
+the initial trace; primary completed/checked it after investigator capacity ended.
+No original physics routine was executed for these assertions.
+
+## Separate owners and gates
+
+47E531 invokes475CA0 before the frame's physics pass47E554→479BC0.475CA0 reads
+5099AB and4F641C only: return `!5099AB ||4F641C`. It does not read controlled
+pointer5099E4. The viewport additionally requires a supervisor and supervisor+8C
+clear; valid QPC dt remains a separate frame condition (loop-motion.md).
+
+Route game complete168 resolves44E450 on LV1B/VR04/LEV1. It stores its argument
+to5099AB at44E458 and clears4F641C at44E465. Jimmy PostLoad calls168(0), while
+modal callers use this state separately. Control118 instead resolves475A40 and
+writes5099E4. Clearing control is not equivalent to calling168 to suspend play.
+
+479BC0 tests body-list enable4F69E7 at479BE9, walks registered payloads rooted
+695D3C and calls479D20 at479C2B. No raw-owner predicate selects that traversal.
+479D58 calls body210→4719F0, reading interface13E; false skips integration.
+Generic construction calls471030(1) at46FDDF..46FDE3, storing authored byte16
+and invoking214→471A00 to set live13E. Init470612/47061B copies authored16
+through getterB4→471020 and setter214. This is the integration gate, not the
+similarly named GAM `CanMove` descriptor used by animated-object behavior.
+
+When eligible, the body consumes pending impulse479EBB..479EF2, acceleration
+479F01 and velocity479F8D regardless of Jimmy's raw ownership.43A750 camera
+preparation clears speed/velocity XYZ but does not clear queued impulse or
+acceleration; the earlier isolated body probe covers that narrow write contract.
+CAM/MCA control transfer does not disable13E in the inspected paths.
+
+## Implementation boundary
+
+The port now runs its existing player body adapter during ordinary cutscene
+control locks, skips controlled movement/input, and still ticks owned fire.
+F1, practice, inspection and inactive-window policy remain whole-frame boundaries.
+This does not reconstruct all supervisor/override/body-list modes, contact masks,
+registration cleanup, static/dynamic response or source cadence. Source support
+flags can differ from the provisional capsule's grounded flag. Other actor bodies
+still use their existing subsystem adapters; this is not a full-scene scheduler.
+
+`test_uncontrolled_physics.py` checks immutable gates/targets and a deliberately
+separate predicate projection across NULL/Jimmy/other control and disabled gates.
+Projection tests are not original execution. REA/Ghidra independently decompiled
+475CA0 and479BC0; see rea.md and control-phase-implementation.md for Evidence IDs.
diff --git a/docs/checkpoint/uncontrolled-player.md b/docs/checkpoint/uncontrolled-player.md
new file mode 100644
index 0000000..ec64c81
--- /dev/null
+++ b/docs/checkpoint/uncontrolled-player.md
@@ -0,0 +1,122 @@
+# Uncontrolled Jimmy and owned jetpack-fire callbacks
+
+Static C3 continuation,2026-10-26. Original `Neutron.exe` SHA256
+`f5281076e1a2492f2d61f87aae3f4544b6d0f535ebe33f9fbb6f78710fe1aea8`.
+All addresses below are preferred-base EXE VAs; inspected raw offsets are
+VA−400000, independently checked through PE sections by the helper. No original
+code was executed for this report. Original files remain read-only. No runtime
+files were changed by this independent investigation.
+
+## Result
+
+Null control is not a gate on the entire Jimmy object or on his owned fire object.
+Jimmy's movement math has an explicit **identity** gate inside437940. Fire421FB0
+does not read the raw controlled pointer at all: after the common gameplay gate
+and base work, it tests its owner/support/timer and follows the owner. Physics is
+an earlier separate pass; its exact activation gates belong to the companion
+physics investigation. Do not infer physics freeze from the movement return.
+
+## Jimmy caller chain and work before the movement gate
+
+Registry473740 invokes interface slot10→470280, which stores dt and dispatches
+normal3C4→424600 (table4A373C). Source editor/supervisor mode uses3C8 instead;
+this report covers normal gameplay. Jimmy game-interface pointer is complete+C0.
+
+424600 increments global4F8408 by dt before inspecting control. At42460A it
+loads5099E4;424622 bypasses only a cleanup prefix when that pointer is nonnull.
+Raw NULL therefore performs UI/resource cleanup, resets interface1D74 and stops
+saved sounds interface974/97C through47D7A0, assigning−1. It does **not** return
+from the object update. Other actors being controlled skip this NULL-only prefix
+but still fail Jimmy's later identity gate.
+
+42469F then queries support218→458960 regardless of control. On support, gadget0
+uses radius160; ordinary selected gadget uses radius60 at4246DC. The callback
+also has its own menu/modal branch42471F..424751, character timers and class
+callbacks. Those are separate from ownership. At424C34 it calls437940; after
+that returns, it continues through complete1C8→424E80 and position/history/class
+callbacks. Complete1C8 has a raw-NULL menu-hide early return of its own.
+
+Consequently “skip all Jimmy update when uncontrolled” and “call the whole
+controller with no keys” are both broader policies than the source gate.
+
+## Movement437940: exact gate placement
+
+| Order / site | Recovered operation |
+|---|---|
+|437946..43794D|Call475CA0; false jumps straight437C33. This is a common gameplay predicate, not the controlled pointer.475CA0 returns false only when5099AB!=0 and4F641C==0; meanings of these two flags are not assigned here.|
+|437953..437964|Increment interface774 timer before raw ownership comparison.|
+|43796A|Base40E050 update runs before raw ownership comparison. Its own resource/transform/animation-end policies remain separate.|
+|43796F..437A0F|Conditional interface7A8 timer/restore work precedes ownership. This is not the jump-hold7B4 field.|
+|437A1B..437A2B|Compare raw5099E4 with Jimmy's game interface; mismatch jumps437C33, the return epilogue. Normal allocated complete pointer is nonzero, so the source complete-to-interface idiom gives this+C0 interface exactly.|
+|437A31..437A43|Two additional flag guards4F83E0 and4F8181 can skip the same controller tail even when Jimmy is controlled. Their full semantics are outside this bounded audit.|
+|437A49 onward|Read movement mode704, jump-held callback and support; choose ground/air application. Grounded released path437B9E clears jump728, mode704 and turn614. These writes are **after** the ownership gate.|
+|437BC5 /437C1C|Invoke complete134→437F90 air/jetpack math or complete130→437C40 ground math. These are complete-object slots, not same-numbered game-interface getters.|
+|437C2D|Write final speed via game-interface26C only after the controlled tail.|
+
+Uncontrolled movement therefore does not itself consume held jump, clear grounded
+jump/mode/turn, compute new steering/velocity/extra acceleration, or perform the
+final speed publication. Existing body state can still be consumed by the earlier
+physics pass. Control-preparation410's explicit speed/velocity clear is a different
+event and must not be expanded into a per-frame reset of every field.
+
+## Fire421FB0: independent owner and timer
+
+Owned fire uses complete+C0 interface4A3160; its normal3C4 target is421FB0.
+Owner is interface5FC=complete6BC; timer is interface600=complete6C0.
+
+| Site | Recovered operation |
+|---|---|
+|421FB6..421FBD|Same475CA0 gate. False returns4220BC.|
+|421FCA|Base40E050 callback before owner test.|
+|421FCF..421FD7|Read owner5FC; no owner skips timer/energy/follow.|
+|421FDF..421FE9|Query owner's support218; true sets remaining time0. Jimmy's support218 resolves458960, which accepts force-support interface4A0 or reads cached support13F via471A10.|
+|421FF3..422003|Subtract dt and store timer, including when already zero/negative.|
+|422009..422049|Strict positive result takes active visibility/energy path; nonpositive hides through complete58(1). Positive path may request complete58(0), then calls42A920 with−dt×.5. The complete70 visibility-related check is not a control gate.|
+|42204C..4220B6|Copy owner logical position310 and Euler328 into fire314/32C even when timer is inactive/hidden.|
+
+Energy helper42A920 skips numeric VR01..VR08, otherwise adjusts global4F83D4 and
+clamps0..100; no controlled-pointer condition. A running timer can continue with
+zero energy, and landing zeroes it before subtracting dt. A controller that is
+skipped while raw NULL does not make the effect skip as well. The full421FB0
+body checksum and absence of the5099E4 address are guarded by the helper; base
+resource/animation work is not replaced with a claim that all indirect operations
+are independent of every global state.
+
+## Collision-side cancellation still runs outside movement ownership
+
+Before calling437940,424600 reads vertical velocity and, if positive, invokes
+complete1A4→425170. This source ceiling-ray/response path is outside raw ownership
+gates. On a hit it can change body velocity and invoke complete1E0→425120. In
+mode1,425120 clears jump complete7E8, requests `FALL`, zeros owned fire6C0 if
+present, then invokes complete140. Thus fire can be cancelled by collision during
+a cutscene even though ownership alone does not cancel it. Complete1A8→4252E0
+also refreshes a source distance cache before437940. These are not yet a recovered
+full collision implementation, and must not be silently replaced by calling
+controlled air math during a cutscene.
+
+After movement, complete1B0→425490 can maintain/replay a position history under
+its own flag; complete1B4→4255C0 has another flagged class-specific effect. Their
+full behavior remains outside this movement/effect split. A limited implementation
+must state that it does not reproduce every branch of424600.
+
+## Port consequence and verification boundary
+
+The audited port gates its entire LabSimulation.Step on effective control, so it
+freezes both body and fire when uncontrolled. A justified partial correction is
+an explicit body/effect tick while suppressing Jimmy input/controller math, with
+control preparation applied at the actual CAM/MCA events. Do not send default
+input through StepPlayer: the grounded-release and steering changes above would
+run despite the source early return. Do not treat raw NULL as F1/practice pause.
+Modern force-control/noclip policy stays distinguishable from original ownership.
+
+Four Python tests verify47 instruction sites,12 vtable targets and two whole-body
+hashes, plus source-gate/state projections with nondefault timers, absent owner,
+raw NULL/other actor, strict expiry and VR energy immunity. Run
+`python3 -m unittest discover -s tests -p test_uncontrolled_player.py -v` or
+`python3 tools/binary/uncontrolled_player.py ../Original` for reproducible JSON.
+These tests interpret immutable code; they do not execute the original callback,
+validate original cadence, prove full body response or compare complete cutscenes.
+
+Remaining: full class-specific424600 branches, pre-control base/timer semantics,
+source collision-side ceiling/distance behavior, complete animation-end timing,
+global475CA0 flag owners and original end-to-end runtime comparison.
diff --git a/docs/compatibility.md b/docs/compatibility.md
index fa03445..ad5ef27 100644
--- a/docs/compatibility.md
+++ b/docs/compatibility.md
@@ -3,6 +3,23 @@
Compilation, native runtime, original behavior and full feature completion are
separate evidence categories. **The game is not fully reconstructed/playable yet.**
+Control-phase continuation (2026-10-26): body/fire eligibility and ordered raw
+preparations are source-byte guarded separately from port behavior. Targeted24
+movement groups, coupled director/preparation tests and the airborne control
+fixture pass. Native Linux GTX1070 at1280x720 passes Carl/control-phase, F1 pause/
+Force Control and F3/tools/same-window checkpoint checks, all with enabled Vulkan
+validation and zero errors. The input-only route uses the actual binding and
+Down braking at turns; source score/task/inventory/contact checks still pass.
+Detailed gaps: checkpoint/control-phase-implementation.md. Native Windows and
+original whole-game comparisons remain unverified for this chunk.
+Full scripts/test.sh passes in artifacts/uncontrolled-phases/integration-final.log:
+93 Python checks,24 movement groups plus remaining managed groups,27 zero-error
+GPU renders with enabled validation, SDL dummy checks and539 unchanged hashes.
+REA6.8.0/Ghidra12.1.4 decompiled five selected procedures locally using a generated
+copy; complete Evidence is retained.421FB0 lacked a recognized procedure; that
+limit is recorded without replacing its existing byte/vtable evidence. Static
+pseudocode is not original execution. See rea.md for local installation and CLI.
+
Retained-state continuation (2026-10-26):14 new source-evidence tests and22
movement groups cover named property loading, cleanup/Init/PostLoad/control
boundaries and bounded route scalar carry. Corrected full tutorial/lab gadget
diff --git a/docs/decisions.md b/docs/decisions.md
index 6e6e368..272661e 100644
--- a/docs/decisions.md
+++ b/docs/decisions.md
@@ -151,3 +151,17 @@
practice gadget after the shared disable pass; LV1B disables1/7. Correct the
implementation and expectations from the later dispatch/call evidence and
retain the earlier report as a scoped historical audit with a correction.
+
+31. **Control identity gates movement, not the whole player.** Run eligible body
+ physics and owned fire while raw control is absent; skip input/controller
+ tail. Preserve queued work and distinct resets. Apply source-specific ordered
+ preparations at CAM/MCA/AI activation; initial MCA target-none returns without
+ completion, and later shots do not repeat preparation. STOP notification follows
+ body clear. Authored transforms update the real player body. Keep modern
+ Force/Noclip and frame pauses separate from raw ownership. Other wrapper/class
+ callbacks and collision remain evidence gates rather than empty-input guesses.
+32. **Use local REA as an additional static evidence source.** Keep Ghidra/JDK
+ and process environment local, analyze generated copies, retain complete
+ Evidence/profile/hash/snapshots and close owned sessions. CLI avoids changing
+ the existing MCP registration. Unknown procedure definitions remain unknown;
+ they do not invalidate byte/vtable evidence or imply missing game code.
diff --git a/docs/progress.md b/docs/progress.md
index 587dcb0..f4f3ea9 100644
--- a/docs/progress.md
+++ b/docs/progress.md
@@ -1,5 +1,54 @@
# Progress — 2026-10-26
+## Uncontrolled-player implementation and REA
+
+Continued C3 into source control/body/fire boundaries. Original475CA0/frame/body
+eligibility is separate from controlled pointer5099E4. Movement437940 has an
+identity gate after base/timer work;421FB0 fire has no raw-control gate. CAM,
+initial MCA and AI preparation order differ; later MCA shots do not repeat it.
+Source-byte reports/helpers/tests: uncontrolled-player.md, uncontrolled-physics.md
+and cutscene-preparation.md under checkpoint/. Investigators supplied independent
+player/control traces; primary completed physics and preparation work when agent
+capacity ended. All file ownership is now primary.
+
+Implemented LabSimulation controlled-input/controller gates while existing body
+physics and owned fire continue. Preparation clears speed/velocity only; queued
+impulse/acceleration and mode/jump/hold/fire are independent. Ordinary grounded
+radius60 is outside the movement gate. PlayerSceneBinding connects ordered raw
+preparations and authored AI root/yaw; STOP observation follows physical clear.
+An initial MCA with first target none returns instead of executing completion.
+Force Control stays an effective override; preparations use authored ownership.
+Headless story pre-roll uses the same phases; F1/practice/inspect still pause them.
+Full wrapper/base transforms/ceiling response, other gadget/class effects and
+registry/cadence remain explicit gaps in control-phase-implementation.md.
+
+Targeted movement24 groups, coupled director tests and the32-frame airborne
+control fixture pass. Native Linux GTX1070 at1280x720 passes Carl/control phases,
+F1 menu/Force Control and F3/tools/checkpoint reload, with zero enabled Vulkan
+validation errors. The input-only lab replay now uses the actual binding; its
+old coasting turn missed the narrow exit. Down braking before large turns reaches
+outside with score150, one practice No, owned jetpack, zero jumps and6 wall hits.
+No physical response/task gate was changed for that replay. A corner snapshot
+is retained for C4. Logs: artifacts/uncontrolled-phases/.
+
+REA MCP connected but lacked native providers. Installed the verified official
+Ghidra12.1.4 ZIP locally and extracted matching full Ubuntu JDK21 packages under
+.tools (system Java21 was runtime-only). Existing Node24/public REA6.8.0 launcher
+is selected by ignored .tools/rea.env and scripts/rea.sh. Client registration was
+not changed. Local Unix IPC needed the approved execution allowance. Five source
+procedures were decompiled and byte-cross-checked; complete Evidence/snapshots are
+on disk.421FB0 maps correctly but has no recognized Ghidra procedure in this
+profile; existing byte/vtable evidence remains authoritative. See docs/rea.md and
+the control-phase report's Evidence IDs. No original game execution or new native
+Windows runtime comparison is claimed by this setup.
+
+Final C3 validation: full scripts/test.sh passes (integration-final.log):93
+Python checks, all managed groups including24 movement, updated input-only route,
+27 GPU render completions with zero enabled Vulkan validation errors, SDL dummy
+lifecycle and539 unchanged originals. Coupled director tests include body-before-
+STOP observation and the initial empty-MCA return. Four native window commands
+pass (five scene renders) in native/*.log. No native Windows runtime is added.
+
## Retained-state implementation continuation
Read prompt.MD in full and resumed C3/C7. Three independent investigators owned
diff --git a/docs/rea.md b/docs/rea.md
new file mode 100644
index 0000000..65f39d3
--- /dev/null
+++ b/docs/rea.md
@@ -0,0 +1,59 @@
+# Local REA and Ghidra
+
+REA is an analysis dependency, absent from the packaged game's runtime. The
+installed MCP tools connected successfully and opened a generated EXE copy with
+the authoritative hash. Native decompilation initially lacked an engine. The
+project-local CLI setup now works without changing the client's MCP registration.
+
+- REA6.8.0: existing public `scripts/rea.mjs` launcher in the user's npm cache.
+- Existing Node24.22.1: configured through ignored `.tools/rea.env`; the login
+ shell's Node20 cannot load this CLI's zstd imports.
+- Ghidra12.1.4: `.tools/ghidra_12.1.4_PUBLIC`, from the
+ [official pinned release](https://github.com/NationalSecurityAgency/ghidra/releases/tag/Ghidra_12.1.4_build).
+ Download ZIP SHA256 `ddac49f903da9d5bac833e5cc79395098b9c33cfd3279be5f31bd00387d2d4db`
+ verified before extraction. Its pinned README requires a full JDK21.
+- Full OpenJDK21.0.12.1+1-1~24.04.4: matching Ubuntu JDK/JRE headless packages
+ downloaded with apt and extracted under `.tools/java`; no system installation.
+ JAVA_HOME is `.tools/java/usr/lib/jvm/java-21-openjdk-amd64`. The pre-existing
+ system Java21 was runtime-only. Adoptium metadata endpoint returned403; no
+ archive from that endpoint was installed.
+
+scripts/rea.sh reads local launcher settings from ignored `.tools/rea.env`, sets
+local GHIDRA_INSTALL_DIR/JAVA_HOME and invokes the public launcher. For another
+machine set NEUTRON_REA_NODE and NEUTRON_REA_LAUNCHER there, or provide `rea` on
+PATH. This environment does not alter the already-running MCP server; native MCP
+tools need those environment paths when that server next starts. CLI use is
+already verified. No registration/setup rewrite was performed.
+
+```bash
+mkdir -p artifacts/rea-copy
+cp ../Original/Neutron.exe artifacts/rea-copy/Neutron.exe
+scripts/rea.sh decompile "$PWD/artifacts/rea-copy/Neutron.exe" 0x475ca0 \
+ --provider ghidra --snapshot "$PWD/artifacts/rea-copy/snapshot.json" \
+ --format json --full-output > artifacts/rea-copy/gate.json
+```
+
+Use generated copies and keep complete Evidence with binary/profile identities.
+For several functions, one local stdio session avoids re-importing the EXE for
+each query:
+
+```bash
+python3 tools/binary/rea_decompile_batch.py artifacts/rea-copy/Neutron.exe \
+ artifacts/rea-batch-NEW 0x40e050 0x437940 0x479bc0
+```
+
+The output directory must be new. The helper saves each complete reply, exports
+successful Evidence and closes/snapshots its owned session. A per-function error
+is retained separately; inspect it before claiming complete coverage. REA's
+Ghidra bridge needs a local Unix socket, which the default Codex sandbox blocked;
+the approved local execution allowance worked. This was a host IPC restriction,
+not an analysis-timeout or engine-install failure. Target executable code is not
+run by these static operations. Retained primitive queries are not a complete
+Ghidra database or proof every function is recognized.
+
+Five selected routines were decompiled and cross-checked against existing byte
+evidence. Fire421FB0 was not recognized as a procedure even with the correct
+file-offset selector; keep that limitation explicit. Actual Evidence IDs and
+gameplay interpretation are in checkpoint/control-phase-implementation.md.
+Logs, readiness errors, complete replies, export bundles and snapshots are under
+artifacts/uncontrolled-phases/rea/. Temporary native sessions were closed.
diff --git a/docs/reverse-engineering.md b/docs/reverse-engineering.md
index 4a05153..3fea4b3 100644
--- a/docs/reverse-engineering.md
+++ b/docs/reverse-engineering.md
@@ -5,6 +5,10 @@ offsets, virtual addresses, API symbols or observed behavior. Decompiled output
evidence, not original source. Original gameplay has not yet been observed successfully; isolated original DLL
probes and modern runtime checks are recorded separately.
+- `checkpoint/control-phase-implementation.md`, `uncontrolled-player.md`,
+ `uncontrolled-physics.md`, `cutscene-preparation.md`: frame/body versus identity
+ gates, ordered CAM/MCA/AI preparation and bounded implementation. `rea.md` records
+ local tooling, Evidence and procedure-recognition limits.
- `checkpoint/retained-player-body.md`, `retained-streamer.md`,
`retained-return-cleanup.md`, `retained-companion.md`: complete named property
read chain, bounded ordinary-load scalar/body resets, full tutorial availability
diff --git a/docs/tasks.md b/docs/tasks.md
index bae98bf..2052a44 100644
--- a/docs/tasks.md
+++ b/docs/tasks.md
@@ -74,9 +74,15 @@ plan in verified chunks. Earlier priorities below are a backlog.
- [x] Full tutorial gadget re-enable tails and LV1B item1/7 restrictions; replaces
previous prefix-only availability. Other level tails/selection remain open.
- [x] Companion/shadow reset trace and equal-state AI guard; no blanket AI carry.
-- [ ] Next: uncontrolled body/controller/fire scheduling, body registration and
- appended-helper callback timing; then C4 support/response and shared viewport/
- DEFAULT handoff. C6 source channel allocation/clock/banks remains separate.
+- [x] Bounded uncontrolled body/fire versus controlled movement, ordered CAM/MCA/
+ AI preparation, no repeated later-shot reset, held input restoration and authored
+ player-transform synchronization; managed/GPU/native Linux checks.
+- [x] Local REA6.8.0/Ghidra12.1.4/full JDK21 launcher, selected native pseudocode
+ Evidence/snapshot export; client MCP configuration unchanged. See rea.md.
+- [ ] Next: full Jimmy/base wrapper and body registration/helper timing; C4 support/
+ response/root and shared viewport/DEFAULT handoff. C6 source channel allocation/
+ clock/banks remains separate. Fire's missing Ghidra procedure definition is an
+ analysis limitation; its immutable source-byte contract is already guarded.
1. Lab regressions are corrected and republished, with user verification of their
own Windows build. Keep comprehensive native Windows driver/audio/input
diff --git a/scripts/rea.sh b/scripts/rea.sh
new file mode 100755
index 0000000..5d3677b
--- /dev/null
+++ b/scripts/rea.sh
@@ -0,0 +1,21 @@
+#!/usr/bin/env bash
+# Local REA launcher. Setup paths live in ignored .tools/rea.env; no client edits.
+set -euo pipefail
+NEUTRON_REA_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
+cd "$NEUTRON_REA_ROOT"
+if [[ -f .tools/rea.env ]]; then source .tools/rea.env; fi
+export GHIDRA_INSTALL_DIR="${GHIDRA_INSTALL_DIR:-$NEUTRON_REA_ROOT/.tools/ghidra_12.1.4_PUBLIC}"
+export JAVA_HOME="${NEUTRON_REA_JAVA_HOME:-$NEUTRON_REA_ROOT/.tools/java/usr/lib/jvm/java-21-openjdk-amd64}"
+NEUTRON_REA_NODE="${NEUTRON_REA_NODE:-node}"
+if [[ ! -x "$JAVA_HOME/bin/javac" || ! -x "$GHIDRA_INSTALL_DIR/support/analyzeHeadless" ]]; then
+ echo 'REA requires the local Ghidra/full JDK setup; see docs/rea.md.' >&2
+ exit 2
+fi
+if [[ -n "${NEUTRON_REA_LAUNCHER:-}" ]]; then
+ exec "$NEUTRON_REA_NODE" "$NEUTRON_REA_LAUNCHER" "$@"
+elif command -v rea >/dev/null; then
+ exec rea "$@"
+else
+ echo 'Set NEUTRON_REA_LAUNCHER to the installed public scripts/rea.mjs; see docs/rea.md.' >&2
+ exit 2
+fi
diff --git a/scripts/test.sh b/scripts/test.sh
index ee6ab71..870270e 100755
--- a/scripts/test.sh
+++ b/scripts/test.sh
@@ -41,6 +41,7 @@ dotnet run --no-build -- --headless --debug-page systems --width 2560 --height 1
dotnet run --no-build -- --headless --debug-overlay --frames 2 --screenshot artifacts/test-debug-overlay.ppm
dotnet run --no-build -- --headless --story --smoke-lab-route
dotnet run --no-build -- --headless --smoke-carl-cutscene --frames 60 --screenshot artifacts/test-carl-cutscene.ppm
+dotnet run --no-build -- --headless --smoke-control-phases
dotnet run --no-build -- --headless --smoke-objective-panel --story-time 1 --screenshot artifacts/test-objective-visible.ppm
dotnet run --no-build -- --headless --smoke-objective-panel --story-time 9 --screenshot artifacts/test-objective-hidden.ppm
dotnet run --no-build -- --headless --scene artifacts/extracted/VR04/scene.json --level gam/VR04.gam --smoke-jetpack --screenshot artifacts/test-jetpack.ppm
diff --git a/src/GamePort/Game/LabSimulation.cs b/src/GamePort/Game/LabSimulation.cs
index c0458f1..2270357 100644
--- a/src/GamePort/Game/LabSimulation.cs
+++ b/src/GamePort/Game/LabSimulation.cs
@@ -57,6 +57,9 @@ public sealed class LabSimulation
: jumpState.JumpState == 3 && jumpState.EffectRemaining > 0 ? PlayerPose.Fly
: !forceFall && Velocity.Y > 0 ? PlayerPose.Jump : PlayerPose.Fall;
public int Steps { get; private set; }
+ public int PhysicsSteps { get; private set; }
+ public int ControllerSteps { get; private set; }
+ public int EffectSteps { get; private set; }
public int Jumps { get; private set; }
public int Landings { get; private set; }
public int WallHits { get; private set; }
@@ -103,6 +106,7 @@ public sealed class LabSimulation
Grounded = true;
Noclip = HasNoclipEntry = false;
Steps = Jumps = Landings = WallHits = 0;
+ PhysicsSteps = ControllerSteps = EffectSteps = 0;
Travel = MaximumRise = 0;
CameraEye = DesiredEye().Eye;
}
@@ -144,21 +148,41 @@ public sealed class LabSimulation
return ((float)Math.Sin(radians), (float)Math.Cos(radians));
}
- public void Step(PlayerInput input)
+ public void Step(PlayerInput input, bool controlled = true)
{
if (Noclip) { StepNoclip(input); return; }
//47E554 global body integration precedes473740 registry callbacks.
// This reconstructs the player/input/fire partial order, not the full
// scene registry, original cadence or original collision response.
StepPhysics();
- if (!InputAfterPlayerUpdate) StepInput(input);
- StepPlayer(input);
+ //42469F..4246DC reasserts ordinary grounded radius before ownership
+ //checks. The scooter's separate radius160 policy remains unimplemented.
+ if (Grounded) InteractionRadius = 60;
+ if (controlled && !InputAfterPlayerUpdate) StepInput(input);
+ if (controlled) StepPlayer(input);
var effect = JetpackBurst.StepEffect(jumpState, Grounded, FixedStepClock.Step, Energy, LevelCode);
jumpState = effect.State; Energy = effect.Energy; JetpackEffectVisible = effect.Visible;
- if (InputAfterPlayerUpdate) StepInput(input);
+ EffectSteps++;
+ if (controlled && InputAfterPlayerUpdate) StepInput(input);
Steps++;
}
+ /// 43A750 preparation clears speed and velocity XYZ only. Queued
+ /// impulse, acceleration, mode, jump/hold and owned fire timer survive.
+ public void PrepareForScript()
+ {
+ Speed = 0; Velocity = Vector3.Zero;
+ }
+
+ /// Authored AI root/Euler writes. They do not reset body motion or
+ /// destination yaw61C; the full logical/render Euler model remains separate.
+ public void ApplyAuthoredTransform(Vector3 root, float yaw)
+ {
+ if (!float.IsFinite(root.X) || !float.IsFinite(root.Y) || !float.IsFinite(root.Z) || !float.IsFinite(yaw))
+ throw new ArgumentOutOfRangeException(nameof(root));
+ Position = root - Vector3.UnitY * rootHeight; Yaw = yaw;
+ }
+
/// Modern debug movement; never serialized into original save state.
public bool SetNoclip(bool enabled)
{
@@ -239,6 +263,7 @@ public sealed class LabSimulation
private void StepPlayer(PlayerInput input)
{
+ ControllerSteps++;
const float dt = FixedStepClock.Step;
var yawError = MathF.IEEERemainder(targetYaw - Yaw, 360);
Yaw = (Yaw + yawError * dt * 3.7f) % 360;
@@ -249,7 +274,6 @@ public sealed class LabSimulation
{
movementMode = 0; TurnRate = 0;
}
- if (Grounded) InteractionRadius = 60;
if (Velocity.LengthSquared() != 0 || movementMode == 0 && lateralContribution != 0)
Yaw = (Yaw + TurnRate * dt) % 360;
var (sin, cos) = Trig(Yaw);
@@ -277,6 +301,7 @@ public sealed class LabSimulation
private void StepPhysics()
{
+ PhysicsSteps++;
const float dt = FixedStepClock.Step;
var wasGrounded = Grounded;
var velocity = Velocity;
@@ -338,5 +363,5 @@ public sealed class LabSimulation
step is >= 80 and < 120, step is >= 60 and < 95, false, step < 180);
public bool SmokePassed => Steps >= 240 && Travel > 50 && MaximumRise > 20 && Jumps > 0 && Landings > 0 && Grounded;
- public string Report => $"Walk checkpoint: steps={Steps}, travel={Travel:F2}, jumps={Jumps}, landings={Landings}, max rise={MaximumRise:F2}, wall hits={WallHits}, grounded={Grounded}, feet={Position}.";
+ public string Report => $"Walk checkpoint: steps={Steps}, controller={ControllerSteps}, physics={PhysicsSteps}, effect={EffectSteps}, travel={Travel:F2}, jumps={Jumps}, landings={Landings}, max rise={MaximumRise:F2}, wall hits={WallHits}, grounded={Grounded}, feet={Position}.";
}
diff --git a/src/GamePort/Game/PlayerSceneBinding.cs b/src/GamePort/Game/PlayerSceneBinding.cs
new file mode 100644
index 0000000..77d0572
--- /dev/null
+++ b/src/GamePort/Game/PlayerSceneBinding.cs
@@ -0,0 +1,56 @@
+using System.Numerics;
+
+namespace GamePort.Game;
+
+/// Connect recovered raw-owner preparations and authored transforms to
+/// the port's player body. Full scene scheduling/other actor classes stay separate.
+public sealed class PlayerSceneBinding : IDisposable
+{
+ private readonly SceneDirector director;
+ private readonly LabSimulation player;
+ private readonly SceneInventory inventory;
+ private readonly string tag;
+ public int UnsupportedGadgetCleanups { get; private set; }
+
+ public PlayerSceneBinding(SceneDirector director, LabSimulation player, SceneInventory inventory, string tag)
+ {
+ this.director = director; this.player = player; this.inventory = inventory; this.tag = tag;
+ director.ActorPreparing += Prepare;
+ director.AuthoredTransformChanged += Transform;
+ }
+
+ public bool Step(PlayerInput input)
+ {
+ var acceptsInput = director.Controlled || player.Noclip;
+ player.Step(input, acceptsInput);
+ var rotation = director.FindActor(tag)?.Rotation ?? Vector3.Zero;
+ director.SyncActor(tag, player.ActorRoot, rotation with { Y = player.Yaw });
+ return acceptsInput;
+ }
+
+ private void Prepare(DirectorActor actor, ScriptPreparation action)
+ {
+ if (!string.Equals(actor.Tag, tag, StringComparison.OrdinalIgnoreCase)) return;
+ if (action == ScriptPreparation.StopMotion) player.PrepareForScript();
+ else if (action == ScriptPreparation.ReleaseGadget)
+ {
+ //427FF0 has distinct cases. Only no-selection/jetpack cleanup is
+ //implemented; do not turn the other gadget cases into blanket resets.
+ if (inventory.SelectedGadget is -1 or 3)
+ { inventory.SelectGadget(-1); player.GadgetMode = -1; }
+ else UnsupportedGadgetCleanups++;
+ }
+ }
+
+ private void Transform(DirectorActor actor)
+ {
+ if (string.Equals(actor.Tag, tag, StringComparison.OrdinalIgnoreCase))
+ player.ApplyAuthoredTransform(actor.Position, actor.Rotation.Y);
+ }
+
+ public void Dispose()
+ {
+ director.ActorPreparing -= Prepare;
+ director.AuthoredTransformChanged -= Transform;
+ }
+}
diff --git a/src/GamePort/Game/SceneDirector.cs b/src/GamePort/Game/SceneDirector.cs
index 5fe54d6..718fe35 100644
--- a/src/GamePort/Game/SceneDirector.cs
+++ b/src/GamePort/Game/SceneDirector.cs
@@ -37,6 +37,7 @@ public sealed class DirectorActor
}
public sealed record DirectorIssue(string Tag, string Kind, string Detail);
+public enum ScriptPreparation { StopMotion, ReleaseGadget, FlattenOrientation }
///
/// Recovered synchronous AI-trigger dispatch and audio-duration-driven CAM/MCA
@@ -71,6 +72,7 @@ public sealed class SceneDirector
public List Actors { get; } = [];
public IReadOnlyList Issues => issues;
private string? controlledTag;
+ public string? AuthoredControlledTag => controlledTag;
/// Modern debug override; leaves authored control/camera transitions intact.
public bool DebugForceControl { get; set; }
public bool Controlled => Equal(ControlledTag, level.Player?.Tag);
@@ -90,6 +92,8 @@ public sealed class SceneDirector
public event Action? MusicStopped;
public event Action? MusicRestarted;
public event Action? ActorChanged;
+ public event Action? ActorPreparing;
+ public event Action? AuthoredTransformChanged;
public event Action? Unsupported;
public SceneDirector(GamLevel level, TaskStates tasks, Func soundDuration)
@@ -277,7 +281,7 @@ public sealed class SceneDirector
var number = activations.GetValueOrDefault(entity.Offset) + 1;
activations[entity.Offset] = number;
if (times != -1 && number > times) return;
- SetControl(entity);
+ SetControl(entity, prepareOwner: true);
var targetTag = entity.Text("AITarget");
if (!IsNone(targetTag))
@@ -286,12 +290,13 @@ public sealed class SceneDirector
Report(entity.Tag!, entity.Kind, $"AI target {targetTag} absent");
else
{
+ var transformed = false;
var hide = entity.Integer("AIHideObj", -1);
if (hide != -1) target.Visible = hide == 0;
var positionTag = entity.Text("AINewPos");
if (!IsNone(positionTag))
{
- if (actors.TryGetValue(positionTag!, out var marker)) target.Position = marker.Position;
+ if (actors.TryGetValue(positionTag!, out var marker)) { target.Position = marker.Position; transformed = true; }
else Report(entity.Tag!, entity.Kind, $"AINewPos {positionTag} absent");
}
var animation = entity.Text("AIAnim");
@@ -302,7 +307,8 @@ public sealed class SceneDirector
var patrol = entity.Text("AIPatrol");
if (!IsNone(patrol)) target.Patrol = patrol;
var yaw = entity.Number("AINewRotY", -1);
- if (yaw != -1) target.Rotation = target.Rotation with { Y = yaw };
+ if (yaw != -1) { target.Rotation = target.Rotation with { Y = yaw }; transformed = true; }
+ if (transformed) AuthoredTransformChanged?.Invoke(target);
ActorChanged?.Invoke(target);
if (target.AIstate == 3 && (target.Kind is "3DOG" or "3CAR") && !patrolConsumers.Contains(target.Tag))
Report(target.Tag, target.Kind, "Patrol motion/arrival is not yet reconstructed; actor-state callback emitted");
@@ -331,15 +337,26 @@ public sealed class SceneDirector
private void StartMulti(GamEntity entity)
{
// MCA's PlayerControlled is its completion target (4311F0), not an
- // activation grant. Each valid shot clears control at43159E..4315A7.
+ // activation grant. Only initial activation clears control; later shots
+ //use431750 without repeating431502..4315A7 preparation.
Toggle(entity);
+ if (IsNone(entity.Text("CameraTarget0"))) return; //431300→431668 returns, not completion.
if (skipping) { FinishMulti(entity); return; }
+ //431502 initial activation only. Subsequent431750 shots do not repeat
+ //preparation or clear control. The raw owner guard ignores debug force.
+ PrepareOwner(ScriptPreparation.StopMotion, ScriptPreparation.FlattenOrientation, ScriptPreparation.ReleaseGadget);
+ ControlledTag = null;
StartCamera(entity, 0, applyCommon: false);
}
private void StartCamera(GamEntity entity, int index, bool applyCommon = true)
{
- if (applyCommon) { SetControl(entity); Toggle(entity); }
+ if (applyCommon)
+ {
+ if (Equal(entity.Text("PlayerControlled"), "NULL"))
+ PrepareOwner(ScriptPreparation.ReleaseGadget, ScriptPreparation.StopMotion, ScriptPreparation.FlattenOrientation);
+ SetControl(entity); Toggle(entity);
+ }
if (entity.Integer("DeactivateInv") != 0) InventoryEnabled = false;
if (skipping)
{
@@ -361,7 +378,6 @@ public sealed class SceneDirector
Block(entity, $"Camera target {targetTag} absent", entity.Text("NextTrigger"));
return;
}
- if (index >= 0) ControlledTag = null;
var database = entity.Text("SoundDatabase") ?? "none";
var duration = soundDuration(database, sound);
if (!double.IsFinite(duration) || duration < 0)
@@ -478,11 +494,33 @@ public sealed class SceneDirector
return target.Position + new Vector3(x * cy - pitchedZ * sy, pitchedY, x * sy + pitchedZ * cy);
}
- private void SetControl(GamEntity entity)
+ private void PrepareOwner(params ScriptPreparation[] actions)
+ {
+ if (controlledTag == null || !actors.TryGetValue(controlledTag, out var owner)) return;
+ foreach (var action in actions)
+ {
+ // Only Jimmy's derived preparation is modeled here; other classes
+ //receive the boundary without inventing their virtual side effects.
+ if (action == ScriptPreparation.StopMotion && owner.Kind == "3JIM")
+ owner.Speed = 0;
+ else if (action == ScriptPreparation.FlattenOrientation)
+ owner.Rotation = new Vector3(0, owner.Rotation.Y, 0);
+ ActorPreparing?.Invoke(owner, action);
+ //42AFC0 invokes43A750 before the derived STOP animation request.
+ if (action == ScriptPreparation.StopMotion && owner.Kind == "3JIM")
+ { owner.Animation = "STOP"; ActorChanged?.Invoke(owner); }
+ }
+ }
+
+ private void SetControl(GamEntity entity, bool prepareOwner = false)
{
var control = entity.Text("PlayerControlled");
if (IsNone(control)) return;
- if (Equal(control, "NULL")) ControlledTag = null;
+ if (Equal(control, "NULL"))
+ {
+ if (prepareOwner) PrepareOwner(ScriptPreparation.StopMotion); //40C4AD; no438/flatten.
+ ControlledTag = null;
+ }
else if (actors.ContainsKey(control!)) ControlledTag = control;
else Report(entity.Tag!, entity.Kind, $"PlayerControlled target {control} absent");
}
diff --git a/src/GamePort/Program.cs b/src/GamePort/Program.cs
index 5780dff..8a9cc24 100644
--- a/src/GamePort/Program.cs
+++ b/src/GamePort/Program.cs
@@ -87,6 +87,7 @@ internal static class Program
--smoke-jetpack 300-step powered steering/jump check in VR04
--smoke-jump-key NAME Replay using shift/right-shift/insert/keypad0/right-mouse
--smoke-carl-cutscene Verify input stays locked during the lab conversation
+ --smoke-control-phases Airborne cutscene body/fire and restoration fixture
--smoke-objective-panel Show original lab objective; --story-time tests timeout
--smoke-objective-window SDL O-key reopen/close/timeout check
--debug-menu Start with the F1 debug menu open
@@ -251,6 +252,7 @@ internal static class Program
var walkSmoke = args.Contains("--smoke-walk");
var jetpackSmoke = args.Contains("--smoke-jetpack");
var carlSmoke = args.Contains("--smoke-carl-cutscene");
+ var controlPhaseSmoke = args.Contains("--smoke-control-phases");
var objectiveWindowSmoke = args.Contains("--smoke-objective-window");
var objectiveSmoke = args.Contains("--smoke-objective-panel") || objectiveWindowSmoke;
var jumpBinding = Option(args, "--smoke-jump-key") ?? "shift";
@@ -268,7 +270,7 @@ internal static class Program
var defaultStory = (defaultScene || args.Contains("--walk")) && !args.Contains("--inspect") && !walkSmoke && !args.Contains("--smoke-window")
&& animationPath == null && !args.Contains("--time");
var story = defaultStory || loadedSave != null || saveOutput != null || args.Contains("--story") || args.Contains("--story-time")
- || args.Contains("--skip-intro") || jetpackSmoke || practiceWindowSmoke || practiceReturnSmoke || carlSmoke || objectiveSmoke;
+ || args.Contains("--skip-intro") || jetpackSmoke || practiceWindowSmoke || practiceReturnSmoke || carlSmoke || controlPhaseSmoke || objectiveSmoke;
if (story && (args.Contains("--inspect") || walkSmoke || args.Contains("--smoke-window")))
throw new ArgumentException("Story playback requires its own simulation mode.");
if (walkSmoke && args.Contains("--inspect")) throw new ArgumentException("Walk smoke cannot run in inspection mode.");
@@ -428,7 +430,9 @@ internal static class Program
var storyTime = double.Parse(Option(args, "--story-time") ?? "0", System.Globalization.CultureInfo.InvariantCulture);
if (!double.IsFinite(storyTime) || storyTime < 0 || storyTime > 3600 || storyTime > 0 && !headless)
throw new ArgumentException("Story time must be0..3600 seconds in headless mode.");
- var frames = int.Parse(Option(args, "--frames") ?? (walkSmoke ? "240" : jetpackSmoke ? "300" : objectiveWindowSmoke ? "180" : carlSmoke ? "60" : toolsSmoke ? "20" : debugSmoke ? "16" : diagnosticsSmoke ? "24" : headless ? "1" : "0"));
+ var frames = int.Parse(Option(args, "--frames") ?? (walkSmoke ? "240" : jetpackSmoke ? "300" : objectiveWindowSmoke ? "180" : carlSmoke ? "60" : controlPhaseSmoke ? "32" : toolsSmoke ? "20" : debugSmoke ? "16" : diagnosticsSmoke ? "24" : headless ? "1" : "0"));
+ if (controlPhaseSmoke && (level?.Code != "LV1B" || loadedSave != null || frames < 32 || director == null))
+ throw new ArgumentException("Control-phase fixture needs a fresh lab and at least32 frames");
if (toolsSmoke && (headless || frames < 16 || director == null || level?.Code != "LV1B" || loadedSave != null || debugMenu.Open))
throw new ArgumentException("Debug tools smoke requires a fresh windowed lab and at least16 frames");
if (debugSmoke && (headless || frames < 16 || director == null || level?.Code != "LV1B" || loadedSave != null || debugMenu.Open))
@@ -450,6 +454,8 @@ internal static class Program
var width = uint.Parse(Option(args, "--width") ?? "960");
var height = uint.Parse(Option(args, "--height") ?? "720");
if (width is < 64 or > 8192 || height is < 64 or > 8192) throw new ArgumentException("Viewport dimensions must be64..8192");
+ using var playerBinding = director == null ? null
+ : new PlayerSceneBinding(director, simulation!, interactions!.Inventory, level!.Player!.Tag!);
var renderer = presentation.LoadScene(mesh, texture, headless,
walking ? WalkTitle : animations?.WindowTitle ?? "Jimmy Neutron — scene preview | C camera",
width, height, !args.Contains("--no-validation"));
@@ -490,7 +496,15 @@ internal static class Program
Console.WriteLine($"Practice RETURN: exact saved root={restored}, root={simulation.ActorRoot}, STRT yaw={simulation.Yaw}, intro suppressed={!director.Playing}, companion restored={companionRestored}, retained player input order={simulation.InputAfterPlayerUpdate}, gadget availability={interactions.Inventory.State(0, 3)}.");
if (!restored) throw new InvalidOperationException("RETURN did not restore the live departure root without replaying STRT actions");
}
- if (args.Contains("--skip-intro") || jetpackSmoke || debugSmoke || diagnosticsSmoke || toolsSmoke) { sceneAudio?.StopSpeech(); director.Skip(); }
+ if (args.Contains("--skip-intro") || jetpackSmoke || debugSmoke || diagnosticsSmoke || toolsSmoke || controlPhaseSmoke) { sceneAudio?.StopSpeech(); director.Skip(); }
+ if (controlPhaseSmoke)
+ {
+ interactions.Inventory.SetState(0, 3, 1); interactions.Inventory.SelectGadget(3);
+ simulation.ApplyRetainedLevelState(new("LV1B", 3, 1, .625f, 6.125f));
+ simulation.RestoreReturnRoot(simulation.ActorRoot + Vector3.UnitY * 200);
+ director.SyncActor(level.Player!.Tag!, simulation.ActorRoot, new(0, simulation.Yaw, 0));
+ director.Trigger("neutron1a");
+ }
if (practiceReturnSmoke && level.Code == "VR04")
{
if (interactions!.Inventory.State(0, 3) != 1)
@@ -540,6 +554,8 @@ internal static class Program
}
for (var tick = 0; tick < (int)Math.Round(storyTime * 60) && !practice.Active && !debugMenu.Open; tick++)
{
+ playerBinding!.Step(default);
+ interactions!.UpdateEnergy(simulation!.Energy);
foreach (var patrol in patrols.Values) patrol.Step(FixedStepClock.Step);
mechanisms!.Step(FixedStepClock.Step);
director.Step(FixedStepClock.Step);
@@ -595,6 +611,8 @@ internal static class Program
var debugPausedSteps = simulation?.Steps ?? 0;
var toolsPaused = false; var toolsMoved = false; var toolsStopped = false; var toolsSaved = false; var toolsSystems = false;
var toolsEntry = simulation?.ActorRoot ?? Vector3.Zero;
+ var controlEntry = simulation?.ActorRoot ?? Vector3.Zero;
+ var phasesLocked = false; var phasesResumed = false;
void ResetScene()
{
if (simulation == null) { debugMenu.Message = "RESET REQUIRES PLAY MODE"; return; }
@@ -846,28 +864,29 @@ internal static class Program
if (debugMenu.Open) animationStarted += elapsed;
if (simulation != null && !debugMenu.Open && !inspecting && !practice.Active && (active || headless || walkSmoke || jetpackSmoke || objectiveWindowSmoke || debugSmoke || diagnosticsSmoke || toolsSmoke))
{
- var steps = headless || walkSmoke || jetpackSmoke || practiceWindowSmoke || carlSmoke || objectiveWindowSmoke || debugSmoke || diagnosticsSmoke || toolsSmoke ? 1 : stepClock.Advance(elapsed);
+ var steps = headless || walkSmoke || jetpackSmoke || practiceWindowSmoke || carlSmoke || controlPhaseSmoke || objectiveWindowSmoke || debugSmoke || diagnosticsSmoke || toolsSmoke ? 1 : stepClock.Advance(elapsed);
for (var step = 0; step < steps; step++)
{
- var input = carlSmoke ? new PlayerInput(true, false, true, false, true)
+ var input = carlSmoke || controlPhaseSmoke ? new PlayerInput(true, false, true, false, true)
: jetpackSmoke && headless ? LabSimulation.JetpackSmokeInput(simulation.Steps)
: walkSmoke && headless ? LabSimulation.SmokeInput(simulation.Steps) : new PlayerInput(
keys.Contains(82) || simulation.Noclip && keys.Contains(26), keys.Contains(81) || simulation.Noclip && keys.Contains(22), keys.Contains(80), keys.Contains(79),
mouseJump || keys.Contains(225) || keys.Contains(229) || keys.Contains(73) || keys.Contains(98),
keys.Contains(44) || keys.Contains(8), keys.Contains(224) || keys.Contains(228) || keys.Contains(20),
keys.Contains(4), keys.Contains(7), keys.Contains(225) || keys.Contains(229));
- if (director?.Controlled != false || simulation.Noclip)
+ var before = simulation.Position;
+ var powered = simulation.JetpackEffectVisible && !simulation.Grounded;
+ bool acceptsInput;
+ if (playerBinding != null) acceptsInput = playerBinding.Step(input);
+ else { simulation.Step(input); acceptsInput = true; }
+ if (jetpackSmoke && powered)
+ {
+ poweredAirTravel += Vector2.Distance(new(before.X, before.Z), new(simulation.Position.X, simulation.Position.Z));
+ poweredAirTurn |= MathF.Abs(simulation.TurnRate) > 1;
+ }
+ interactions?.UpdateEnergy(simulation.Energy);
+ if (acceptsInput)
{
- var before = simulation.Position;
- var powered = simulation.JetpackEffectVisible && !simulation.Grounded;
- simulation.Step(input);
- if (jetpackSmoke && powered)
- {
- poweredAirTravel += Vector2.Distance(new(before.X, before.Z), new(simulation.Position.X, simulation.Position.Z));
- poweredAirTurn |= MathF.Abs(simulation.TurnRate) > 1;
- }
- interactions?.UpdateEnergy(simulation.Energy);
- director?.SyncActor(level!.Player!.Tag!, simulation.ActorRoot, new Vector3(0, simulation.Yaw, 0));
if (!simulation.Noclip)
{
director?.Touch(level!.Player!.Tag!, simulation.ActorRoot);
@@ -976,6 +995,16 @@ internal static class Program
diagnostics.ObserveFrame(presentedAt - lastPresentedAt);
lastPresentedAt = presentedAt;
count++;
+ if (controlPhaseSmoke && count == 15)
+ {
+ phasesLocked = director is { Controlled: false, Playing: true }
+ && simulation is { ControllerSteps: 0, PhysicsSteps: 15, EffectSteps: 15, Jumps: 0, Speed: 0, TurnRate: 0 }
+ && simulation.ActorRoot.Y < controlEntry.Y && simulation.JetpackRemaining is > 0 and < 6.125f
+ && simulation.Energy < 20 && Vector2.Distance(new(simulation.ActorRoot.X, simulation.ActorRoot.Z), new(controlEntry.X, controlEntry.Z)) < .001f;
+ director!.Skip();
+ }
+ if (controlPhaseSmoke && count == 30)
+ phasesResumed = director!.Controlled && simulation!.ControllerSteps == 15 && simulation.Travel > 1;
audio?.Pump();
sceneAudio?.Pump();
if (toolsSmoke)
@@ -1198,10 +1227,17 @@ internal static class Program
}
if (carlSmoke)
{
- var locked = director is { Controlled: false, Playing: true, CurrentTag: "neutron1a" } && simulation!.Steps == 0;
- Console.WriteLine($"Carl cutscene input lock with held move/turn/jump: {locked}.");
+ var locked = director is { Controlled: false, Playing: true, CurrentTag: "neutron1a" }
+ && simulation is { ControllerSteps: 0, Jumps: 0, Speed: 0, TurnRate: 0 }
+ && simulation.PhysicsSteps == simulation.Steps && simulation.EffectSteps == simulation.Steps && simulation.Steps >= frames;
+ Console.WriteLine($"Carl cutscene input lock with held move/turn/jump: {locked}; body/fire continue={simulation!.PhysicsSteps}/{simulation.EffectSteps}.");
if (!locked) return 2;
}
+ if (controlPhaseSmoke)
+ {
+ Console.WriteLine($"Control-phase fixture: locked input with body/fire/energy continuing={phasesLocked}; held-input resumption={phasesResumed}.");
+ if (!phasesLocked || !phasesResumed) return 2;
+ }
if (objectiveSmoke) Console.WriteLine($"Objective panel: time={objectivePanel!.Elapsed:F3}, pad={objectivePanel.PadVisible}, content={objectivePanel.ContentVisible}, offset={objectivePanel.YOffset:F3}.");
if (objectiveWindowSmoke)
{
diff --git a/tests/GamePort.DirectorTests/Program.cs b/tests/GamePort.DirectorTests/Program.cs
index afc6cdf..fe88886 100644
--- a/tests/GamePort.DirectorTests/Program.cs
+++ b/tests/GamePort.DirectorTests/Program.cs
@@ -139,6 +139,96 @@ Check(carlBytes.AsSpan(0x3159e, 15).SequenceEqual(Convert.FromHexString("8B0D489
"Source MCA clears game controlled pointer before playback");
Console.WriteLine("PASS end-of-lab Carl cutscene input lock and natural/skip control restoration");
+foreach (var kind in new[] { "3CAM", "3MCA", "3AIT" })
+{
+ var jim = Entity("3JIM", "JIM1", ("RotationX", 12f), ("RotationZ", 22f));
+ var command = Entity(kind, "command", ("PlayerControlled", "NULL"),
+ ("CameraTarget", "JIM1"), ("CameraTarget0", "JIM1"), ("CameraTarget1", "JIM1"),
+ ("SoundIndex", 1), ("SoundIndex0", 1), ("SoundIndex1", 1), ("SoundDatabase", "voice"),
+ ("NextTrigger", "default"));
+ var scene = new SceneDirector(new("LV1B", [jim, command]), new TaskStates([]), (_, _) => .1);
+ var body = ControlBody();
+ var bag = new SceneInventory(new int[19]); bag.SetState(0, 3, 1); bag.SelectGadget(3);
+ using var binding = new PlayerSceneBinding(scene, body, bag, "JIM1");
+ for (var tick = 0; tick < 5; tick++) binding.Step(new(true, false, false, false, false));
+ var callbacks = new List();
+ scene.ActorPreparing += (actor, action) =>
+ {
+ Check(actor.Tag == "JIM1" && scene.AuthoredControlledTag == "JIM1", "Preparation observes the prior raw owner");
+ callbacks.Add(action);
+ };
+ var stopObserved = false;
+ scene.ActorChanged += actor =>
+ {
+ if (actor.Tag == "JIM1" && actor.Animation == "STOP")
+ {
+ Check(body.Velocity == Vector3.Zero && body.Speed == 0, "Body preparation precedes derived STOP notification");
+ stopObserved = true;
+ }
+ };
+ scene.Trigger("command");
+ var expected = kind switch
+ {
+ "3CAM" => new[] { ScriptPreparation.ReleaseGadget, ScriptPreparation.StopMotion, ScriptPreparation.FlattenOrientation },
+ "3MCA" => new[] { ScriptPreparation.StopMotion, ScriptPreparation.FlattenOrientation, ScriptPreparation.ReleaseGadget },
+ _ => new[] { ScriptPreparation.StopMotion }
+ };
+ Check(stopObserved && callbacks.SequenceEqual(expected) && body.Speed == 0 && body.Velocity == Vector3.Zero,
+ "CAM/MCA/AI retain their own ordered preparation boundaries");
+ Check(bag.SelectedGadget == (kind == "3AIT" ? 3 : -1), "AI control loss does not imply camera gadget cleanup");
+ Check(scene.FindActor("JIM1")!.Rotation.X == (kind == "3AIT" ? 12 : 0), "Only camera preparation flattens pitch/roll");
+ if (kind == "3AIT") continue; // Its DEFAULT next trigger has already restored control.
+ var controllers = body.ControllerSteps;
+ binding.Step(new(true, false, true, false, true));
+ Check(body.ControllerSteps == controllers && body.Jumps == 0 && body.PhysicsSteps > controllers,
+ "Input lock leaves body/effect callbacks running");
+ scene.DebugForceControl = true;
+ binding.Step(new(true, false, false, false, false));
+ var speed = body.Speed;
+ scene.Step(.11f);
+ Check(callbacks.SequenceEqual(expected), "Later MCA shots/CAM completion do not replay initial preparations");
+ Check(body.Speed == speed && scene.Controlled, "Force control does not become the raw preparation owner");
+ scene.DebugForceControl = false;
+ scene.Trigger("default");
+ var jumps = body.Jumps;
+ binding.Step(new(false, false, false, false, true));
+ Check(body.Jumps == jumps + 1, "DEFAULT restores held jump without an invented fresh-key boundary");
+}
+Console.WriteLine("PASS ordered CAM/MCA/AI preparations, body continuation, raw-owner force override and held-input restoration");
+
+{
+ var jim = Entity("3JIM", "JIM1");
+ var command = Entity("3AIT", "move", ("PlayerControlled", "NULL"), ("AITarget", "JIM1"),
+ ("AINewPos", "marker"), ("AINewRotY", 75f));
+ var marker = Entity("DUMY", "marker", ("PositionX", 200f), ("PositionY", 500f), ("PositionZ", 300f));
+ var scene = new SceneDirector(new("LV1B", [jim, command, marker]), new TaskStates([]), (_, _) => 1);
+ var body = ControlBody();
+ using var binding = new PlayerSceneBinding(scene, body, new SceneInventory(new int[19]), "JIM1");
+ scene.Trigger("move");
+ Near(body.ActorRoot, marker.Position, "Authored AI position is applied to the actual player body");
+ Check(body.Yaw == 75 && !scene.Controlled, "Authored AI yaw follows the logical setter during control loss");
+ binding.Step(new(true, false, true, false, true));
+ Check(body.ActorRoot.Y < marker.Position.Y && body.ActorRoot.X == marker.Position.X && body.ControllerSteps == 0,
+ "Body settles at the scripted location without accepting movement input");
+ Near(scene.FindActor("JIM1")!.Position, body.ActorRoot, "Physics publishes the new logical player root");
+}
+Console.WriteLine("PASS authored player teleport/yaw survives uncontrolled body synchronization");
+
+{
+ var jim = Entity("3JIM", "JIM1");
+ var empty = Entity("3MCA", "empty", ("CameraTarget0", "none"), ("PlayerControlled", "NULL"), ("NextTrigger", "default"));
+ var missing = Entity("3CAM", "missing", ("CameraTarget", "absent"), ("SoundIndex", 1), ("PlayerControlled", "NULL"));
+ var scene = new SceneDirector(new("LV1B", [jim, empty, missing]), new TaskStates([]), (_, _) => 1);
+ var callbacks = new List(); scene.ActorPreparing += (_, action) => callbacks.Add(action);
+ scene.Trigger("empty");
+ Check(callbacks.Count == 0 && scene.AuthoredControlledTag == "JIM1" && !scene.Playing,
+ "Initial MCA none target returns without preparation or invoking completion ownership");
+ scene.Trigger("missing");
+ Check(callbacks.Count == 3 && scene.AuthoredControlledTag == null && scene.Blocked,
+ "CAM preparation precedes target-lookup failure");
+}
+Console.WriteLine("PASS first-MCA none return and CAM failure preparation boundaries");
+
foreach(var type in new[]{0,1,2,3,4})
{
var camera=Entity("3MCA","camera",("CameraTarget0","actor"),("SoundDatabase","voice"),("SoundIndex0",1),("CameraType0",type),("LookatVOffset0",100f),("NextTrigger","default"));
@@ -268,6 +358,15 @@ var returnGate=vr.Entities.First(e=>e.Kind=="LOAD");interactions.Update(returnGa
Check(transitions==1 && returned!.Returning && returned.LevelFile==save.ReturnLevel && returned.StartPoint==save.ReturnStart,"RETURN uses stored destination, restore flag and contact-enter gating");
Console.WriteLine("PASS original VR01 star/trophy score, strict contact entry, acquired-state save/reload and RETURN request");
+static LabSimulation ControlBody()
+{
+ var points = new[] { new Vector3(-10000, 0, -10000), new Vector3(-10000, 0, 10000),
+ new Vector3(10000, 0, 10000), new Vector3(10000, 0, -10000) };
+ var floor = new Mesh(points.Select(p => new MeshVertex(p, Vector2.Zero, Vector3.One)).ToArray(), [0, 1, 2, 0, 2, 3]);
+ var standing = new Mesh([new(new(-20, -20, -20), Vector2.Zero, Vector3.One), new(new(20, 80, 20), Vector2.Zero, Vector3.One)], []);
+ return new(new SceneCollision(floor), standing, new(0, 25, 0), 0);
+}
+
static GamEntity Entity(string kind,string tag,params (string Name,object Value)[] values)
{
var properties=new Dictionary();
diff --git a/tests/GamePort.LabRouteTests/Program.cs b/tests/GamePort.LabRouteTests/Program.cs
index bf01f10..ecf8c84 100644
--- a/tests/GamePort.LabRouteTests/Program.cs
+++ b/tests/GamePort.LabRouteTests/Program.cs
@@ -41,6 +41,7 @@ interactions.PracticeRequested += practice.Show;
var prompts = 0;
LevelTransition? destination = null;
interactions.Transition += request => destination ??= request;
+using var playerBinding = new PlayerSceneBinding(director, simulation, interactions.Inventory, level.Player!.Tag!);
director.Begin(start);
interactions.Reset(); mechanisms.Reset(); director.Skip();
@@ -70,12 +71,18 @@ for (var leg = 0; leg < route.Length && destination == null; leg++)
if (difference.Length() < 90) { arrived = true; break; }
var desiredYaw = MathF.Atan2(-difference.X, difference.Y) * 180 / MathF.PI;
var error = MathF.IEEERemainder(desiredYaw - simulation.Yaw, 360);
- if (director.Controlled)
+ var previous = simulation.ActorRoot;
+ // This is a modern input driver, not original navigation. Brake before
+ //large heading changes; coasting turns can miss the narrow exit opening
+ //when the recovered cutscene preparation changes the arrival speed.
+ var brake = MathF.Abs(error) > 15 && simulation.Speed > 0;
+ var input = brake ? new PlayerInput(false, true, false, false, false)
+ : new PlayerInput(MathF.Abs(error) < 10, false, error > 1, error < -1, false);
+ var acceptsInput = playerBinding.Step(input);
+ Require(Vector3.Distance(previous, simulation.ActorRoot) < 100, "Player warped during input replay");
+ interactions.UpdateEnergy(simulation.Energy);
+ if (acceptsInput)
{
- var previous = simulation.ActorRoot;
- simulation.Step(new(MathF.Abs(error) < 30, false, error > 1, error < -1, false));
- Require(Vector3.Distance(previous, simulation.ActorRoot) < 100, "Player warped during input replay");
- director.SyncActor(level.Player!.Tag!, simulation.ActorRoot, new(0, simulation.Yaw, 0));
director.Touch(level.Player!.Tag!, simulation.ActorRoot);
interactions.Update(simulation.ActorRoot, simulation.InteractionRadius);
if (practice.Active) continue; // Same modal activation boundary as Program.
@@ -86,7 +93,7 @@ for (var leg = 0; leg < route.Length && destination == null; leg++)
director.Step(FixedStepClock.Step);
if (destination != null) break;
}
- Require(arrived || destination != null, $"Input route blocked at leg{leg}: {simulation.ActorRoot}");
+ Require(arrived || destination != null, $"Input route blocked at leg{leg}: {simulation.Report}; yaw={simulation.Yaw}, control={director.AuthoredControlledTag}, shot={director.CurrentTag}, blocked={director.Blocked}, goal={route[leg]}");
Require(tasks.Get("SCENE") == sceneStates[leg], $"Wrong task state after walking leg{leg}");
Console.WriteLine($"PASS walked leg{leg}: SCENE={tasks.Get("SCENE")}, root={simulation.ActorRoot}");
}
diff --git a/tests/GamePort.MovementTests/Program.cs b/tests/GamePort.MovementTests/Program.cs
index 3162d70..2b794c5 100644
--- a/tests/GamePort.MovementTests/Program.cs
+++ b/tests/GamePort.MovementTests/Program.cs
@@ -10,6 +10,8 @@ var tests = new (string Name, Action Check)[]
("clock remainder, stall bound, reset and invalid intervals", ClockBoundaries),
("physics consumes prior motion; fresh versus retained input order", FrameOrdering),
("ordinary route load retains scalars independently of body/jump reset", RetainedLevelState),
+ ("script preparation preserves queued work and uncontrolled physics consumes it", UncontrolledPhysics),
+ ("uncontrolled fire ticks while movement and grounded-release reset are gated", UncontrolledFire),
("source camera query origin, hit adjustment, gain and one query per input", CameraQueryConsumer),
("debug noclip bypasses walls/gravity/camera and exits without queued motion", Noclip),
("turn-only speed, forward acceleration/cap and release", SpeedLimits),
@@ -103,6 +105,60 @@ static void FrameOrdering()
Check(jump.MovementMode == 0, "reset restores stored walking input mode");
}
+static void UncontrolledPhysics()
+{
+ var cameraQueries = 0;
+ var simulation = FlatSimulation(cameraRay: (_, _) => { cameraQueries++; return null; });
+ simulation.Step(new(true, false, false, false, true));
+ var pending = simulation.PendingImpulse;
+ var mode = simulation.MovementMode; var jump = simulation.JumpState;
+ var hold = simulation.JumpHoldSeconds; var fire = simulation.JetpackRemaining;
+ var before = simulation.Position;
+ simulation.PrepareForScript();
+ Check(simulation.Speed == 0 && simulation.Velocity == Vector3.Zero && simulation.PendingImpulse == pending,
+ "43A750 zeros speed/velocity while queued impulse survives");
+ Check(simulation.MovementMode == mode && simulation.JumpState == jump && simulation.JumpHoldSeconds == hold
+ && simulation.JetpackRemaining == fire, "Preparation does not invent a mode/jump/hold/fire reset");
+ var queries = cameraQueries;
+ for (var tick = 0; tick < 8; tick++) simulation.Step(new(true, false, true, false, true), controlled: false);
+ Check(simulation.Position.Y > before.Y && simulation.PendingImpulse == Vector3.Zero,
+ "Uncontrolled body consumes the queued impulse once and keeps integrating");
+ Check(simulation.Jumps == 1 && simulation.ControllerSteps == 1 && simulation.Speed == 0 && simulation.TurnRate == 0
+ && simulation.JumpState == jump && simulation.MovementMode == mode && simulation.JumpHoldSeconds == hold,
+ "Held movement/turn/jump cannot enter the uncontrolled controller");
+ Check(simulation.PhysicsSteps == 9 && simulation.EffectSteps == 9 && simulation.Steps == 9,
+ "Body and fire callbacks remain distinct from controlled movement count");
+ Check(cameraQueries == queries, "Uncontrolled input does not advance the follow-camera query");
+ simulation.ApplyAuthoredTransform(simulation.ActorRoot + new Vector3(25, 0, 30), 85);
+ Check(simulation.Velocity.Y > 0 && simulation.Jumps == 1,
+ "Authored transform write is separate from body preparation");
+}
+
+static void UncontrolledFire()
+{
+ foreach (var code in new[] { "LV1B", "VR04" })
+ {
+ var simulation = FlatSimulation(); simulation.LevelCode = code;
+ simulation.ApplyRetainedLevelState(new("LV1B", 3, 1, .625f, 6.125f));
+ simulation.RestoreReturnRoot(simulation.ActorRoot + Vector3.UnitY * 500);
+ var energy = simulation.Energy;
+ simulation.PrepareForScript();
+ simulation.Step(new(true, false, true, false, true), controlled: false);
+ Near(simulation.JetpackRemaining, 6.125f - FixedStepClock.Step, .00001f, "Owned fire time continues during control lock");
+ Near(simulation.Energy, code == "VR04" ? energy : energy - FixedStepClock.Step * .5f, .00001f,
+ "Fire drain and tutorial immunity do not depend on raw control");
+ Check(simulation.ControllerSteps == 0 && simulation.JumpState == 0 && simulation.MovementMode == 1
+ && simulation.JumpHoldSeconds == .625f && simulation.JetpackEffectVisible, "Controller scalars untouched while fire remains visible");
+ }
+ var grounded = FlatSimulation();
+ grounded.ApplyRetainedLevelState(new("LV1B", 3, 1, .625f, 6.125f));
+ grounded.Step(default, controlled: false);
+ Check(grounded.Grounded && grounded.MovementMode == 1 && grounded.JetpackRemaining < 0
+ && grounded.Energy == 20, "Grounded fire clears independently of the gated grounded-release controller");
+ grounded.Step(default);
+ Check(grounded.MovementMode == 0 && grounded.ControllerSteps == 1, "Normal grounded-release reset resumes with control");
+}
+
static void RetainedLevelState()
{
var state = new PlayerLevelState("LV1B", 3, 1, .625f, 6.125f);
diff --git a/tests/test_cutscene_preparation.py b/tests/test_cutscene_preparation.py
new file mode 100644
index 0000000..ba7fed0
--- /dev/null
+++ b/tests/test_cutscene_preparation.py
@@ -0,0 +1,36 @@
+from dataclasses import replace
+from pathlib import Path
+import sys
+import unittest
+
+ROOT = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(ROOT / 'tools/binary'))
+from cutscene_preparation import collect, preparation_events, JetpackPreparationState, project_jetpack_preparation
+
+
+class CutscenePreparationTests(unittest.TestCase):
+ def test_source_bytes_and_virtual_targets(self):
+ self.assertEqual(collect(ROOT.parent / 'Original')['sha256'],
+ 'f5281076e1a2492f2d61f87aae3f4544b6d0f535ebe33f9fbb6f78710fe1aea8')
+
+ def test_distinct_preparation_order_and_later_shot_boundary(self):
+ self.assertEqual(preparation_events('cam_null', 'JIM1'),
+ ['JIM1:438', 'JIM1:410', 'JIM1:flatten_euler', 'control:NULL'])
+ self.assertEqual(preparation_events('mca_activation', 'JIM1'),
+ ['JIM1:410', 'JIM1:flatten_euler', 'JIM1:438', 'control:NULL'])
+ self.assertEqual(preparation_events('ai_null', 'JIM1'), ['JIM1:410', 'control:NULL'])
+ self.assertEqual(preparation_events('mca_next_shot', 'JIM1'), [])
+ self.assertEqual(preparation_events('default_restore', 'JIM1'), [])
+ self.assertEqual(preparation_events('mca_activation', None), [])
+
+ def test_bounded_jetpack_state_does_not_become_blanket_reset(self):
+ before = JetpackPreparationState()
+ self.assertEqual(project_jetpack_preparation(before, 'cam_null'),
+ replace(before, speed=0, velocity=(0, 0, 0), selected_gadget=-1, euler=(0, 150, 0)))
+ self.assertEqual(project_jetpack_preparation(before, 'ai_null'), replace(before, speed=0, velocity=(0, 0, 0)))
+ with self.assertRaises(ValueError):
+ project_jetpack_preparation(replace(before, selected_gadget=0), 'cam_null')
+
+
+if __name__ == '__main__':
+ unittest.main()
diff --git a/tests/test_uncontrolled_physics.py b/tests/test_uncontrolled_physics.py
new file mode 100644
index 0000000..0096d1b
--- /dev/null
+++ b/tests/test_uncontrolled_physics.py
@@ -0,0 +1,25 @@
+from pathlib import Path
+import sys
+import unittest
+
+ROOT = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(ROOT / 'tools/binary'))
+from uncontrolled_physics import collect, eligible
+
+
+class UncontrolledPhysicsTests(unittest.TestCase):
+ def test_source_field_and_virtual_gate_provenance(self):
+ self.assertEqual(len(collect(ROOT.parent / 'Original')['tables']), 6)
+
+ def test_raw_control_does_not_replace_frame_and_body_flags(self):
+ for control in (None, 0x1000, 0x2000):
+ self.assertTrue(eligible(raw_control=control))
+ for gate in ('frame_valid', 'supervisor', 'bodies_enabled', 'registered', 'body_integration'):
+ self.assertFalse(eligible(raw_control=control, **{gate: False}), gate)
+ self.assertFalse(eligible(raw_control=control, editor=True))
+ self.assertFalse(eligible(raw_control=control, flag5099ab=True))
+ self.assertTrue(eligible(raw_control=control, flag5099ab=True, override4f641c=True))
+
+
+if __name__ == '__main__':
+ unittest.main()
diff --git a/tests/test_uncontrolled_player.py b/tests/test_uncontrolled_player.py
new file mode 100644
index 0000000..8540c8d
--- /dev/null
+++ b/tests/test_uncontrolled_player.py
@@ -0,0 +1,53 @@
+"""Original ownership-gate evidence and independent controller/effect boundaries."""
+from pathlib import Path
+import sys
+import unittest
+
+ROOT = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(ROOT / 'tools/binary'))
+from uncontrolled_player import collect, controller_phases, fire_projection
+
+
+class UncontrolledPlayerTests(unittest.TestCase):
+ @classmethod
+ def setUpClass(cls):
+ cls.report = collect(ROOT.parent / 'Original')
+
+ def test_source_instruction_virtual_and_whole_body_provenance(self):
+ self.assertEqual(len(self.report['sites']), 47)
+ bodies = {x['name']: x for x in self.report['bodies']}
+ self.assertFalse(bodies['complete_fire_callback']['raw_control_reference'])
+ self.assertTrue(bodies['complete_movement_callback']['raw_control_reference'])
+
+ def test_null_and_other_control_skip_only_gated_controller_tail(self):
+ for raw in (None, 0, 0x2000):
+ phases = controller_phases(raw)
+ self.assertIn('base40E050', phases)
+ self.assertIn('conditional_timer7A8', phases)
+ self.assertNotIn('controller_modes', phases)
+ self.assertNotIn('jump_ground_air', phases)
+ self.assertIn('jump_ground_air', controller_phases(0x1000))
+ self.assertNotIn('controller_modes', controller_phases(0x1000, status83e0=True))
+ self.assertNotIn('controller_modes', controller_phases(0x1000, modal8181=True))
+ self.assertEqual(controller_phases(0x1000, game_enabled=False), ['game_gate'])
+
+ def test_fire_timer_is_not_controller_or_cutscene_ownership_state(self):
+ # The exact same owned-fire tick can follow any controller ownership
+ # projection: raw control is not an input to the original fire callback.
+ for raw in (None, 0x1000, 0x2000):
+ controller_phases(raw)
+ self.assertEqual(fire_projection(2, .25, 20, 0x4C563142), (1.75, 19.875, True, True))
+ self.assertEqual(fire_projection(.25, .25, 20, 0x4C563142), (0, 20, False, True))
+ self.assertEqual(fire_projection(2, .25, 20, 0x4C563142, grounded=True), (-.25, 20, False, True))
+ self.assertEqual(fire_projection(-1, .25, 20, 0x4C563142), (-1.25, 20, False, True))
+
+ def test_fire_gate_owner_and_tutorial_energy_are_independent_boundaries(self):
+ self.assertEqual(fire_projection(2, .25, 20, 0x4C563142, owner=False), (2, 20, False, False))
+ self.assertEqual(fire_projection(2, .25, 20, 0x4C563142, game_enabled=False), (2, 20, False, False))
+ for code in range(0x56523031, 0x56523039):
+ self.assertEqual(fire_projection(2, .25, 20, code), (1.75, 20, True, True))
+ self.assertEqual(fire_projection(2, .25, .01, 0x4C563142), (1.75, 0, True, True))
+
+
+if __name__ == '__main__':
+ unittest.main()
diff --git a/tools/binary/cutscene_preparation.py b/tools/binary/cutscene_preparation.py
new file mode 100644
index 0000000..cb3658c
--- /dev/null
+++ b/tools/binary/cutscene_preparation.py
@@ -0,0 +1,179 @@
+#!/usr/bin/env python3
+"""Read-only CAM/MCA/AI preparation provenance and bounded event projection."""
+from __future__ import annotations
+
+import argparse
+from dataclasses import dataclass, replace
+import hashlib
+import json
+from pathlib import Path
+import struct
+
+from checkpoint_script_state import EXE_SHA256, find_case, read_va
+from pe_inspect import PE
+
+
+SITES = (
+ ('cam_already_active_guard', 0x4165AB, '389e800900000f856c040000'),
+ ('cam_control_none_branch', 0x4165F8, '0f84fc000000'),
+ ('cam_control_null_compare', 0x4165FE, '6864d04e00'),
+ ('cam_named_control', 0x41662A, 'ff9218010000'),
+ ('cam_raw_owner_guard', 0x416635, 'a1e49950003bc30f84a9000000'),
+ ('cam_owner438', 0x41665D, 'ff9038040000'),
+ ('cam_owner410', 0x41666B, 'ff9210040000'),
+ ('cam_owner_euler', 0x41667E, 'ff9028030000'),
+ ('cam_cached_euler', 0x4166A0, '81c174020000'),
+ ('cam_flatten_logical_euler', 0x4166E5, 'ff922c030000'),
+ ('cam_clear_control', 0x4166F4, 'ff9018010000'),
+ ('cam_target_resolution_after_control', 0x416754, 'e817d90500'),
+ ('mca_activation_index_reset', 0x4312A7, '899ea40e0000'),
+ ('mca_first_none_guard', 0x431300, '0f8462030000'),
+ ('mca_first_none_return', 0x431668, '5f5e5d5b83c420c3'),
+ ('mca_first_target_lookup', 0x43131A, 'e8512d0400'),
+ ('mca_first_target_missing_continues', 0x431330, '0f84e9000000'),
+ ('mca_target_animation_prelude', 0x431417, 'ff92e0000000'),
+ ('mca_raw_owner_guard', 0x431502, '8b0de49950003bcb0f849d000000'),
+ ('mca_owner410', 0x431512, 'ff9210040000'),
+ ('mca_owner_euler', 0x431525, 'ff9028030000'),
+ ('mca_flatten_logical_euler', 0x431565, 'ff922c030000'),
+ ('mca_cached_euler', 0x431578, '0574020000'),
+ ('mca_owner438', 0x431598, 'ff9038040000'),
+ ('mca_clear_control', 0x4315A7, 'ff9218010000'),
+ ('mca_start_first_shot', 0x431600, 'ff90e0000000'),
+ ('mca_later_shot_dispatch', 0x430E87, 'ff90e0000000'),
+ ('mca_finish_dispatch', 0x430ED8, 'ff92e4000000'),
+ ('mca_shot_sound_guard', 0x431771, '83bc86e40d0000ff'),
+ ('mca_shot_target_lookup', 0x4317C1, 'e8aa280400'),
+ ('mca_shot_sound_start', 0x431838, 'e853bb0400'),
+ ('mca_shot_target_animation', 0x431909, 'ff92e0000000'),
+ ('mca_completion_null_compare', 0x4311F4, '6864d04e00'),
+ ('mca_completion_control', 0x431222, 'ff9218010000'),
+ ('ai_interface_to_complete', 0x40CA72, '81c138ffffffe883f8ffff'),
+ ('ai_control_descriptor', 0x40C1A5, '8d868c0900006a01506818ce4e00'),
+ ('ai_null_control_guard', 0x40C4AD, '8b0de499500085c97466'),
+ ('ai_owner410', 0x40C4B9, 'ff9210040000'),
+ ('ai_clear_control', 0x40C527, 'ff9218010000'),
+ ('ai_marker_root', 0x40C732, 'ff9010030000'),
+ ('ai_set_root', 0x40C757, 'ff9214030000'),
+ ('ai_yaw_sentinel', 0x40C88F, '81bd940b0000000080bf'),
+ ('ai_set_authored_yaw', 0x40C93E, 'ff9274010000'),
+ ('ai_set_cached_euler', 0x40C94C, '81c334030000'),
+ ('jimmy_prepare_base', 0x42AFC3, 'e888f70000'),
+ ('jimmy_prepare_stop', 0x42AFDE, '6840d04e008bceff9250010000'),
+ ('prepare_speed_zero', 0x43A758, 'c7860806000000000000'),
+ ('prepare_velocity_xyz', 0x43A764, '6a006a006a008bceff90c4020000'),
+ ('jimmy438_cleanup', 0x42AE08, '8b8640ffffff8d8e40ffffffff9078010000'),
+ ('jimmy438_runtime_flags', 0x42AE1A, 'b0018886651d0000a282814f00'),
+ ('default_already_owned', 0x447D5C, '3bc6753f'),
+ ('default_restore_control', 0x447E27, 'ff9018010000'),
+ ('default_snap_strict', 0x447E9D, 'd81dace54800dfe0f6c441'),
+ ('input_owner_guard', 0x475A9C, '8b0de499500085c97409'),
+ ('input_owner_dispatch', 0x475AA9, 'ff92cc030000'),
+ ('jump_held_raw_checks', 0x4378DC, 'a06598500084c0751fa048985000'),
+)
+
+TABLES = (
+ ('CAM activation', 0x498060, 0xD8, 0x4165A0),
+ ('MCA activation', 0x4A93C8, 0xD8, 0x431280),
+ ('MCA per-shot', 0x4A93C8, 0xE0, 0x431750),
+ ('MCA completion', 0x4A93C8, 0xE4, 0x4311F0),
+ ('Jimmy410', 0x4A373C, 0x410, 0x42AFC0),
+ ('Jimmy438', 0x4A373C, 0x438, 0x42AE00),
+ ('Jimmy Euler getter', 0x4A373C, 0x328, 0x4723F0),
+ ('Jimmy Euler setter', 0x4A373C, 0x32C, 0x472450),
+ ('Jimmy position setter', 0x4A373C, 0x314, 0x472360),
+ ('Jimmy authored Euler setter', 0x4A373C, 0x174, 0x4716D0),
+ ('Jimmy stop sound', 0x4A3BDC, 0x148, 0x429A00),
+ ('Jimmy action request', 0x4A3BDC, 0x150, 0x43AFF0),
+ ('Jimmy gadget cleanup', 0x4A3BDC, 0x178, 0x427FF0),
+ ('Jimmy held jump', 0x4A3BDC, 0x158, 0x4378D0),
+)
+
+# Full body identity guards negative claims as well as selected positive sites.
+BODY_RANGES = (('MCA per-shot', 0x431750, 0x4319A2),
+ ('held jump getter', 0x4378D0, 0x43793A))
+
+
+def preparation_events(kind: str, raw_owner: str | None) -> list[str]:
+ """Local source-event projection, not an original routine execution.
+
+ Activation preconditions have already reached the inspected preparation
+ block. Real target, sound and skip handling are intentionally not simulated.
+ """
+ events = []
+ if kind == 'cam_null':
+ if raw_owner:
+ events += [raw_owner + ':438', raw_owner + ':410', raw_owner + ':flatten_euler']
+ events += ['control:NULL']
+ elif kind == 'mca_activation':
+ if raw_owner:
+ events += [raw_owner + ':410', raw_owner + ':flatten_euler', raw_owner + ':438', 'control:NULL']
+ elif kind == 'ai_null':
+ if raw_owner:
+ events += [raw_owner + ':410']
+ events += ['control:NULL']
+ elif kind in ('mca_next_shot', 'default_restore'):
+ pass # Neither boundary dispatches410/438.
+ else:
+ raise ValueError('Unknown preparation boundary')
+ return events
+
+
+@dataclass(frozen=True)
+class JetpackPreparationState:
+ speed: float = 350
+ velocity: tuple[float, float, float] = (10, 20, 30)
+ impulse: tuple[float, float, float] = (1, 2, 3)
+ acceleration: tuple[float, float, float] = (4, 5, 6)
+ mode: int = 1
+ jump: int = 3
+ hold: float = .625
+ fire: float = 6.125
+ selected_gadget: int = 3
+ euler: tuple[float, float, float] = (15, 150, 25)
+
+
+def project_jetpack_preparation(state: JetpackPreparationState, kind: str):
+ """Bounded−1/3 selected-gadget projection; do not generalize other gadgets."""
+ if state.selected_gadget not in (-1, 3):
+ raise ValueError('Other gadget cleanup policies are outside this fixture')
+ result = state
+ for event in preparation_events(kind, 'JIM1'):
+ if event.endswith(':410'):
+ result = replace(result, speed=0, velocity=(0, 0, 0))
+ elif event.endswith(':438'):
+ result = replace(result, selected_gadget=-1)
+ elif event.endswith(':flatten_euler'):
+ result = replace(result, euler=(0, result.euler[1], 0))
+ return result
+
+
+def collect(root):
+ pe = PE(find_case(Path(root), 'Neutron.exe'))
+ if hashlib.sha256(pe.data).hexdigest() != EXE_SHA256:
+ raise ValueError('Unsupported source EXE')
+ sites = []
+ for name, va, expected in SITES:
+ actual = read_va(pe, va, len(bytes.fromhex(expected))).hex()
+ if actual != expected:
+ raise ValueError(f'{name} changed at {va:#x}: {actual}')
+ sites.append(dict(name=name, va=hex(va), file_offset=hex(pe.offset(va-pe.base)), bytes=actual))
+ tables = []
+ for name, table, slot, target in TABLES:
+ actual, = struct.unpack(' 0
+ if active and not 0x56523031 <= level_code <= 0x56523038:
+ energy = max(0, min(100, energy - delta * .5))
+ return remaining, energy, active, True
+
+
+def collect(root):
+ pe = PE(find_case(Path(root), 'Neutron.exe'))
+ if hashlib.sha256(pe.data).hexdigest() != EXE_SHA256:
+ raise ValueError('Unsupported original EXE')
+ sites = []
+ for name, va, expected in SITES:
+ actual = read_va(pe, va, len(bytes.fromhex(expected))).hex()
+ if actual != expected:
+ raise ValueError(f'Changed instruction {name} at {va:#x}: {actual}')
+ sites.append(dict(name=name, va=hex(va), file_offset=hex(pe.offset(va-pe.base)), bytes=actual))
+ tables = []
+ for name, table, slot, target in TABLES:
+ if struct.unpack('