EscapistsReintroduced/docs/current-status.md

# Current development status

**20 September 2026 — nearby sighting handoff and resupply approach:** A squad without a usable own radio route can now pass its fresh visual report through an original member to a living friendly radio carrier within 100 m. Original source, position and timestamp remain intact; HQ learns only after radio delivery. Shared callers cannot overlap their originating calls, and interrupted or repeated sightings cannot bypass delivery. The [native checkpoint](evidence/native-2026-09-20-101306-local-handoff/report.json) completes one round/81 assertions without script/runtime failures: an arranged snapshot reaches HQ unchanged, while removing the second caller's radio prevents the next snapshot from arriving. The fixture deliberately removes radios and positions a witness after startup fitting settles; it is not natural perception or multiplayer proof. [Details and retained setup failure](hq-local-handoff.md). Four external localization errors remain, with source/external fingerprints and hidden/muted cleanup confirmed.

Resupply now waits through unsubmitted geometry/capacity failures instead of spending that magazine and aborting the trip. A submitted native attempt remains strictly once-only. Move/hold radii also no longer accept an approach five metres from the reported truck position. Actual-method regressions reproduce and fix both issues; full native walking rendezvous remains unproved. Full repository checks pass, with the final fixture adjustment covered by focused tests and native compilation/execution. All 533 reference sources remain unchanged. An updated editable source bundle is recorded [here](evidence/source-2026-09-20-handoff-resupply.json); packaged playtests remain older snapshots. Manual play, multiplayer, packaging and FPS qualification are still unfinished.

**20 September 2026 — native finite magazine transfer:** Before moving a magazine, infantry now need unobstructed physical reach to a root-owned authored truck storage context. The [native transfer checkpoint](evidence/native-2026-09-20-091854-infantry-transfer/report.json) completes one round and 43 assertions, with zero script/runtime failures: one original magazine moves from the automatic convoy's original ammo truck to an original patrol actor, preserving ammunition count and the force ledger. This fixture deliberately positions the recipient and waits for its native pose update; it does **not** prove autonomous rendezvous. Four installed external localization errors still prevent strict qualification. [Access and retained diagnostic failures](infantry-resupply-access.md). Actual-method tests cover obstructions, stock identity, report consumption, reservations, native submission and interrupted fixture cleanup. Source/external input fingerprints and hidden/muted cleanup are confirmed. Full walking resupply, multiplayer, FPS and refreshed packaging remain unfinished; existing handoffs are unchanged.

**20 September 2026 — automatic convoy and result feedback:** The full AI scenario now defaults to one automatic initial convoy after a 30-second startup grace and existing-force admission. Automatic and manual requests share the same finite receipt, with no automatic retry or replacement. `/esc hq convoy auto set 0|1` saves the next-round choice; old disabled convoy settings migrate correctly. [Startup details](convoy-startup.md). Original physical turret occupants receive scoped skill100, while other enemy crew retain at least80. The [native checkpoint](evidence/native-2026-09-20-081658-automatic-convoy/report.json) completes 93 assertions and one round, proving six automatic vehicles, 19 original seated crew, five turret specializations and duplicate-dispatch rejection, plus the existing HQ radio-memory tests. Zero script/runtime failures; four external localization errors remain. All repository checks pass, and 533 tracked reference sources remain unchanged. Hidden/muted cleanup and unchanged input fingerprints are confirmed. Defeat screens now refresh a late replicated end reason; callback/lifecycle tests pass, but real remote-client timing remains untested. Existing packaged handoffs remain older snapshots; native logistics, multiplayer and FPS qualification remain unfinished.

**20 September 2026 — finite nearby infantry resupply:** HQ now uses delivered low-ammunition reports to reserve an existing patrol and the manually dispatched convoy's original ammo truck. Separate radio orders stop the truck, request its stopped-position report and send infantry to a nearby rendezvous. Existing captured magazine identities transfer at most once; no replacement stock is created. Low-stock reporting uses 30-second local scans and a 60-second urgent-call cooldown. [Scope and limits](infantry-resupply.md). The [current native run](evidence/native-2026-09-20-072839-infantry-resupply/report.json) compiles the complete integration and completes 59 HQ-memory assertions/one round with zero script/runtime failures and verified hidden/muted cleanup; four external localization errors remain. The first native compile failure is retained and corrected. Direct native resupply/navigation, automatic dispatch, broad logistics parity and multiplayer remain unfinished. All repository checks pass and all 533 tracked reference sources remain unchanged. The existing packaged handoff remains the older `915476f` snapshot.

**20 September 2026 — radio and finite maintenance:** Enemy groups now use bounded one-time radio fitting, preserving existing radios and excluding mortar-owned allocations. Uplinks hold their original caller for three seconds before network hops and fail on live consciousness/ownership loss. The [native radio/HQ-memory checkpoint](evidence/native-2026-09-20-062203-radio-call-memory/report.json) completes 59 assertions and one round, with zero script/runtime failures; four installed external localization errors remain. All six tested groups were already equipped, so native insertion remains unproved. The western supply cache now holds one finite repair kit and two fuel cans. Its [native stock/transport checkpoint](evidence/native-2026-09-20-062011-maintenance-stock/report.json) passes 131 assertions with zero captured errors, confirming 20 supply units and two five-litre cans plus prior lock/vehicle-loss behavior. Ordinary repair/refuel consumption remains untested. Both runs preserve exact sources and confirm hidden/muted cleanup. Current external packaging has a separate [explicit v3 development workflow](packed-external-development.md); the matching separate handoff is `.build/playtest-2026-09-20-radio-maintenance`. All five PAKs match their authored payloads; [Core ordinary startup](evidence/packed-core-2026-09-20-radio-maintenance/report.json) passes without captured errors. Full packed Chat verifies the exact thirteen-addon closure but remains failed on installed localization errors and did not establish participant readiness before its diagnostic ended. Native build shutdown failures remain; older handoffs are unchanged. All repository checks pass and 533 tracked reference files remain unchanged.

**20 September 2026 — optics and base interaction follow-up:** Startup optics fitting now uses verified vanilla/RHS resources, preserves existing attachments and remains independent of finite Stinger processing. The supply UAZ has a successor-owned [held lock-breach action and scoped entry guards](vehicle-security.md); the [native supply-vehicle checkpoint](evidence/native-2026-09-20-052915-supply-vehicle-security/report.json) completes 124 assertions with zero captured errors, including lock rejection, one-time unlocking, native driver availability with its declared door context, scripted seating, fuel depletion and no replacement after loss. Ordinary held input and multiplayer verification remain pending. Player guidance now explains pilot responsibility and whole-round extraction consequences, and repeats the current objective after a controlled-body change. [The native convoy diagnostic](evidence/native-2026-09-20-050906-convoy-seat-diagnostic/report.json) compiled the optics integration but failed a combined gunner consciousness/seat/identity check after 170 passing assertions; it did not reach the new radio HOLD/RESUME test. The [second run](evidence/native-2026-09-20-051505-convoy-cover-diagnostic/report.json) passed those driving/seat checks but received no qualifying cover sighting before a combined preservation guard failed. Native HOLD/RESUME remains unverified; more precise diagnostics are included without weakening either guard. The [final full-addon run](evidence/native-2026-09-20-053057-optics-security-hq-integration/report.json) compiles all current source changes and completes 56 HQ-memory assertions with no script/runtime failures; the four known external localization errors remain. All repository checks pass, and all 533 tracked reference sources are unchanged. Existing packaged playtests remain older snapshots.

**20 September 2026 — delivered ammunition and convoy cover:** HQ now receives detached small-arms ammunition counts with real radio status and avoids new armed assignments only when fresh, complete evidence proves the squad empty. Per-vehicle convoy HOLD/RESUME commands share the original crews' radio downlinks and preserve route/crew/cargo identity. The [ammunition run](evidence/native-2026-09-20-040857-hq-ammunition/report.json) completes 49 assertions; the [current convoy driving recheck](evidence/native-2026-09-20-041150-convoy-cover-driving/report.json) completes 181. Both have zero script/runtime failures and confirmed cleanup, with four external localization errors retained. Native combat HOLD/RESUME and depletion still need direct tests; general resupply, multiplayer/FPS qualification and updated packaged playtests remain unfinished. See [ammunition scope](hq-ammunition.md) and [convoy scope](convoy-migration.md#delivered-hq-mounted-cover-follow-up).

**20 September 2026 — delivered casualty follow-up:** HQ now merges body and squad-status evidence by immutable original member identity. A [native corpse/radio test](evidence/native-2026-09-20-030533-body-radio-casualties/report.json) completes45 assertions: actual visual observation, no knowledge before delivery, real overlapping status and no double-counting. Transfer-before-death and deleted-corpse safeguards are covered by focused regressions. Four external localization errors remain; packaged playtest binaries have not been refreshed.

**20 September 2026 — HQ memory and external dependencies:** The editable successor now includes the [AI migration checkpoint](ai-migration-progress.md): bounded report-derived memory/prediction, infantry task selection, delivered negative-search coverage, routine check-ins and missing-patrol investigations, local body observation, finite Stinger grants and the Core force ledger. RHS/WCS/BMP are [declared and audited](external-addons.md). Native memory (56 assertions) and actual-sighting HQ response (35 assertions) complete with zero script/runtime failures; four external localization errors keep strict qualification failed. The older playtest below has not been rebuilt with these changes. The [convoy checkpoint](convoy-migration.md) now proves exact nineteen-person native seating, finite physical truck cargo and duplicate-dispatch rejection. The native driving follow-up completes181 assertions: all six vehicles move at least20m and the lead over100m, preserving exact crew and rejecting repeat dispatch. Combat tactics, unloading and long-route reliability remain unfinished.

**20 September 2026 — refreshed playtest ready:** `.build/playtest-2026-09-20` now contains the mission-owned Chat menu and current owned AI/Intel behavior from source `5e74269`. All 216 authored source files and all five native PAK payloads match; fifteen runtime files are verified. [Core](evidence/packed-core-2026-09-20-playtest/report.json) and [full scenario](evidence/packed-chat-2026-09-20-playtest/report.json) pass ordinary packaged startup with zero captured errors, unchanged inputs and confirmed hidden/muted cleanup. [Build record](evidence/playtest-2026-09-20/build.json) and [launch/route instructions](playtest-build.md). Separate muted manual launchers and local profiles/runtime are prepared; no visible launcher was run. The old handoff and 533 original reference files are unchanged. The packager's shutdown failures remain, while a narrowly fixed disappearing-audio-stream race passes44 tests and its separate Chat retry produces matching artifacts. Human full escape, multiplayer and FPS remain unqualified.

**19 September 2026 — mission-owned chat administration:** Core now owns `/esc`, the native settings/HQ/journal menu, authenticated sessions and player tools in dedicated classes. AI and Intel register optional providers; Chat is only the existing full-scenario/development wrapper. Open with `/esc menu`. [Implementation and command reference](chat-integration.md). [Final native regression](evidence/native-2026-09-19-234618-chat-complete/report.json): one full scripted escape round, 101 assertions, zero captured errors, unchanged source and verified private cleanup; checks include actual menu/owner RPCs, saved settings, god-state restoration, visible journal text and scrolling to the newest entry. Earlier Core and two-round full-scenario runs are linked in the implementation notes. All repository checks pass. Original 533 reference files remain unchanged. The frozen playtest package has not been rebuilt; multiplayer and release acceptance remain pending.

**19 September 2026, 22:59 — owned AI/Intel migration:** Added admin Intel summary/reroll commands, native-operation callbacks and bounded retry handling, coalesced team discovery notifications, and fair per-squad reporting of multiple current visual contacts. [Change details](owned-ai-intel-migration.md). The updated five-addon source passes [two native escape/restart rounds](evidence/native-2026-09-19-225422-owned-ai-intel/report.json), **186 assertions**, zero captured errors, unchanged source and confirmed hidden/muted test cleanup. All repository checks pass. The frozen playtest package below remains its earlier snapshot; these source changes have not been repackaged. Human/multiplayer and release qualification remain pending.

Updated **19 September 2026,22:12**. The prison and the complete intended base sequence are implemented: prisoner spawning, physical intel/scavenging, radio defence, player-flown helicopter extraction, individual results and restart. Five independent successor projects use vanilla assets without the original mod. Scripted local integration and ordinary startup have native evidence. This is an implemented development build awaiting gameplay qualification, not a completed multiplayer release.

**Priority correction from the user:** stop spending disproportionate time on automated prison keypress proof and experimental AI aircraft flight. Both experiments are paused. Active work now targets concrete base gameplay bugs, missing user-visible behavior and companion integration. The reproduced first-admission deadlock is fixed, and casualty loading now gives denial/failure feedback to its initiating rescuer. Each clue cache carries one finite replacement map. Manual play and real multiplayer remain necessary, but absence of automated input proof does not mean the prison is still being built.

The frozen202-file playtest source now passes [two Core escape/restart rounds](evidence/native-2026-09-19-215437-playtest-core-loops/report.json),616 assertions, zero captured errors and confirmed cleanup. All repository checks pass. The [development playtest handoff](playtest-build.md) is ready in `.build/playtest-2026-09-19`: exact-source native packages and editable source, separate user-run `play-full.sh`/`play-core.sh`, prepared isolated local profiles/runtime and a seven-step route guide. [Packed Core](evidence/packed-core-2026-09-19-playtest/report.json) and [full five-addon startup](evidence/packed-chat-2026-09-19-playtest/report.json) both pass with verified cleanup and unchanged artifacts. Workbench packaging errors remain recorded separately; this is a development candidate for human testing, not a qualified release.

Recent retained checkpoints:

- [Ordinary coarse approach/facing/lock selection](evidence/game-2026-09-19-212117-coarse-facing-selection/runner.json.txt) passed using actual private mouse/W input and one wheel pair. Facing took2.726seconds; no F was sent. The [first opt-in held-action attempt](evidence/game-2026-09-19-214138-held-prerequisite-refusal/runner.json.txt) later refused its independent facing prerequisite before wheel/F, confirmed all releases and cleaned up. Further input automation is paused.
- [Typed aircraft service20m route](evidence/native-2026-09-19-210233-typed-service-route/report.json) passed40 assertions, destination hold, contact landing and shutdown at its201-file snapshot. The100m survey experiments found terrain/building/tree obstructions before allocating an aircraft; the [last survey](evidence/native-2026-09-19-213726-field-workshop-refusal/report.json) records the workshop bounds. There is no100m flight or operational HQ aircraft qualification. A reviewed contact check prevents ground friction receiving airborne-braking credit; its focused policy checks pass, native validation remains pending.
- All repository checks passed after integrating the optional held-action helper and independent validators. The subsequent narrow contact fix has focused regressions. These are development checks, not manual gameplay or release acceptance.

**All seventeen first-release acceptance cases remain pending.** Full ordinary escape, real multiplayer, improved FPS and clean native packaging are unqualified. The [empty-addon packaging control](evidence/package-control-2026-09-19-201532-exit-after-init/review.json) still reproduces25 shutdown errors. Tests remain hidden/muted, and the original533 reference files were verified unchanged at21:42.

The [route gain comparison](evidence/route-comparison-2026-09-19-192238/comparison.json) measures one native20m route per gain: translation through arrival hold improved155.365→69.293seconds, and final target error improved0.130→0.099m. The candidate passed38assertions, physical contact landing and engine shutdown. This is an isolated flight-controller result, not an FPS comparison or operational HQ air support; the production callback bridge has separate native evidence, while the later retained-service route/Stop passes are separate historical snapshots and useful field-length routes remain pending.

[Ordinary prison approach191750](evidence/game-2026-09-19-191750-prison-approach/runner.json.txt) now passes end to end using actual private mouse/W input:14turns,9walking pulses,4.741m path,27.199s native interval, stationary0.490m from the interior waypoint. A disposable2560x720 viewport provides pointer travel while retaining every containment/ownership check. No gate action was sent; the exact lock/context remains unselected from the final view. Earlier pointer-edge, cross-position-API, calibration-return and legacy-summary failures remain archived. The first physically completed route191342 has a separate hashed raw-log review, not a rewritten runner verdict. The later [194128 alignment attempt](evidence/alignment-review-2026-09-19-194128/review.json) acquires the native gate context with the lock available, but does not select the lock; it remains failed. Read-only selected-action diagnostics precede any action cycling or heldF proof.

The [relative-look observer](ordinary-look-observer.md) now passes an [ordinary Chat run](evidence/game-2026-09-19-183944-relative-look/runner.json.txt): actual +24/−24 pixel input produced a measured 1.935° yaw response and return within 0.149°, with stationary body, then verified W/Tab input. Zero captured errors, source unchanged and cleanup confirmed. The earlier run remains failed because its independent checker assumed identical SAMPLE/READY milliseconds; its [separate corrected review](evidence/look-review-2026-09-19-183353/review.json) preserves the raw archive. The current-source Core attempt reached 292 assertions without errors but exhausted an undersized 180-second budget; its [timeout](evidence/native-2026-09-19-183622-core-budget/runner.json.txt) remains a failure. The later 184647 run used a 600-second budget and completed both rounds with614 assertions; the earlier timeout remains failed. No gate interaction or full normal-input escape is qualified.

The primary deliverable is an independently authored replacement for the abandoned Escapists base mission, with companion behavior integrated through its interfaces. Five successor addon projects have their own identities, authored Everon sites and vanilla asset references. The first release targets up to twelve players completing prison → scavenging/intel → radio defence → helicopter extraction without the original mod or dependent extensions. Scripted development checkpoints exist; a dependable, manually playable multiplayer release is still unfinished.

## Established development checkpoints

| Area | Evidence | Practical limit |
| --- | --- | --- |
| Integrated base loop | [Core 130347](evidence/native-2026-09-19-130347-core-loops/report.json) completes two rounds, 612 assertions and zero errors, including canonical participant rejection/restored document recovery, rough/exact native markers, radio feedback, extraction and restart. [Companion 131041](evidence/native-2026-09-19-131041-companion-loops/report.json) completes two rounds, 96 assertions and zero errors at its source snapshot. | Scripted local integration with deliberate state/position changes, not a normal-input escape. These earlier archives share their captured source snapshot; the new Core/companion/ordinary runs linked above qualify the later native-end correction. The earlier flight mechanism remains separately source-bound. Future addon edits require fresh qualification. Earlier failures remain archived. |
| Ordinary game startup | [Chat 125906](evidence/game-2026-09-19-125906-integrated-prison/report.json) loads all five addons in the regular game, verifies the header/MissionSystems and admits the real local controller to a living canonical prisoner, then observes a clean ready interval. Native screenshots show the real prisoner view and the correct escape objective; the stock-scale yard decal creates a larger irregular bare patch, with substantial wheat remaining. | Explicit diagnostic noSplash and a scoped standalone round-local identity; no complete ordinary route or remote identity proof. The latest bounded keyboard cycle described above passes with visible inventory, while the earlier failed cycles remain archived. [Admission design](local-singleplayer-admission.md). |
| Packed addon consumption | [Packed Core 110729](evidence/packed-core-2026-09-19-110729/report.json) passes ordinary startup from the archived project/PAK/resource database plus installed vanilla: exactly three expected addons, real canonical prisoner, a 15-second ready interval and zero captured errors. Package and log hashes are retained. | Artifact-bound startup with explicit noSplash; no loose successor addon was loaded. The artifact predates current mortar edits. [Archived packed Chat 131942](evidence/packed-chat-2026-09-19-131942-current/report.json) passes startup from all five [1317 packages](evidence/package-2026-09-19-1317-all/review.json), with no loose successor source, a real prisoner and zero captured errors. All 151 PAK entries match the captured 183-file addon snapshot before the native-end correction; all fifteen runtime artifacts retain exact hashes. Native packaging remains unqualified: an [empty-addon control](native-packaging-control.md) reproduces the same UI resource-leak errors, without fixing or suppressing them. |
| Native game-over | [Preparation 133847](evidence/native-2026-09-19-133847-native-end-preparing/report.json) and [active 134044](evidence/native-2026-09-19-134044-native-end-active/report.json) each pass two rounds/21 assertions: actual inherited game-over, unchanged native data, one shutdown, cancelled preparation/tick/binding, inert stale callbacks and vanilla restart. | Local fixture; pending native preloads and real clients are not arranged. Existing successor extraction results remain covered by normal escape/restart regressions. |
| Prison and scavenging | [Prisoner capacity](evidence/native-2026-09-19-063949/report.json) uses one real player plus eleven physical characters, exact starter kits and native clue transfers. [Prison boundary 100148](evidence/native-2026-09-19-100148/report.json) covers the gate/progression corrections. | Extra physical actors are not twelve connected clients. Native action completion and deliberately repositioned boundary crossings do not prove held input or walking through the gate/debris. Equipment scarcity and combat balance need playtesting. |
| Medicine and extraction casualties | [Medical treatment](evidence/native-2026-09-19-062406/report.json) demonstrates finite supplies and native recovery at its earlier snapshot. [Casualty 102037](evidence/native-2026-09-19-102037/report.json) passes 101 clean assertions at the committed checkpoint: grounded-aircraft checks, real incapacitation, helper rejection, seat/LOS constraints, one owner-handshake load and idempotent stop. [Native removal](evidence/native-2026-09-19-080058/report.json) proves the inherited exit mechanism. | Casualty loading explicitly adapts helper membership and invokes action completion. Held input, real remote ownership, competing rescuers and disconnect/death/deletion races remain separate. The older medical snapshot retains its historical catalog errors. |
| Infantry and HQ | [Actual perception response](evidence/native-2026-09-19-092237/report.json) carries an original soldier's observation through both physical radio legs and HQ to a different original squad, which physically moves. Patrol/preemption/resumption are implemented. | The prisoner is arranged and temporarily protected. Broader combat and interrupted/captured support require additional qualification; no response force is created for this test. |
| Finite support | [Mortar expiry](evidence/native-2026-09-19-091306/report.json) verifies fresh native commitments followed by stale-report withholding without duplicate orders. [Courier 095513](evidence/native-2026-09-19-095513/report.json) passes 154 clean assertions through actual radio, original-patrol walking, exact depot-shell delivery and patrol restoration. | Mortar observations are synthetic; impact accuracy and genuine late callbacks remain unproved. The new refresh now passes the bounded 112900 fixture: three actual radio updates, eight physical launches, native event completion, 19 surviving original shells and 25 seconds without replay. The current [114751 expiry regression](evidence/native-2026-09-19-114751-mortar-expiry/report.json) also passes 48 clean assertions. [115232 crew registration](evidence/native-2026-09-19-115232-intel-readiness/report.json) proves all four physical crew receive their one Intel carrying roll after setup, then exercises inventory depletion/loss; broader support-loss checks remain separate. Courier depletion is arranged by physically sequestering the original 27 rounds, not firing or replacing them. Broader losses and multiplayer remain unqualified. |
| Aircraft mechanism | [Flight 114504](evidence/native-2026-09-19-114504-powered-flight/report.json) passes 81 assertions with zero captured errors: signed actor-local collective consumption, native autohover, powered takeoff, five-second 7–9m hover, stable contact landing and an engine-off check with finite fuel. The same aircraft and living AI pilot remain inside the original physical limits. | [Translation 120245](evidence/native-2026-09-19-120245-powered-translation/report.json) additionally passes 85 clean assertions for a fixed four-metre target, five-second hold and landing. The reusable production controller separately passes [123439](evidence/native-2026-09-19-123439-aircraft-controller/report.json), 37 assertions including actual post-spawn qualification and its own complete vertical operation. [Airborne cancellation 124915](evidence/native-2026-09-19-124915-aircraft-cancel/report.json) passes 39 assertions with retained control through contact landing and engine-off. The same reusable controller then passes [fixed-target movement 130038](evidence/native-2026-09-19-130038-aircraft-target/report.json), 38 assertions including a physical 2.9m translation and landing. These are isolated flight mechanisms; rotors still coast after engine shutdown. [The fixed 20 m route](evidence/native-2026-09-19-155909-aircraft-route/report.json) now passes one unloaded flight, target hold, contact landing and engine-off with 38 assertions. Useful field-length navigation, operational HQ support, repeated-flight reliability and manual/multiplayer piloting remain unqualified. Earlier input, drift, hover and shutdown failures remain archived. |
| Shared field map | [Map ownership 120633](evidence/native-2026-09-19-120633-map-ownership/report.json) passes native replacement-view, delayed-callback, mode-clear and spare/interrupted deletion checks. The fixture observes actual cleanup and waits for native component references to expire. | Camera restore authorization is observed; no public native render-state getter exists. Visible held-map usability, previously focused stock gadget callbacks and remote presentation remain separate. |
| Independent runtime controls | [Archived packed Linux Chat 132456](evidence/dedicated-2026-09-19-132456-packed-chat/review.json) compiles the integrated game scripts, loads all five packages and prepares the empty prison mission in an isolated network namespace. Its four resource errors also occur in the earlier vanilla control. All fifteen packed inputs remain unchanged. [Vanilla ordinary 102533](evidence/vanilla-game-2026-09-19-102533/report.json) reproduces the initial splash unknown-class error with exactly two vanilla addons and no successor. | No real dedicated client joined. The vanilla ordinary control remains failed despite entering the menu; it does not test scenario reload or input. [Splash diagnosis](normal-splash-audit.md). |

## Immediate work

1. Play the frozen development candidate through the normal route and fix reported blockers. First-spawn exhaustion, silent casualty-loading failures and missing spare maps have been addressed; do not resume disproportionate input/flight automation.
2. Review normal-player guidance and supplies across the authored escape route. Fix evidenced usability defects with proportionate checks.
3. Distinguish genuinely absent base behavior from existing behavior awaiting playtesting. Pre-round setup UI, broader island population and vehicle break-in/alarm behavior are potential remaining features; the narrow escape sequence already exists. Keep later-scope feature parity separate.
4. Prepare and validate the playable development build. Ordinary manual play, real multiplayer, packaging and equivalent-workload FPS tests remain release work. Do not resume elaborate gate-input or experimental aircraft automation at the expense of those priorities.

## First-release qualification

The [acceptance ledger](acceptance.json) still requires the complete ordinary escape loop, death/unconsciousness/recovery, reconnect/respawn/spectators, partial escape, defeat/restart, interrupted/stale reports, support/ammunition losses and duplicate-free cleanup with up to twelve connected players. A passed development checkpoint does not promote those pending cases.

Collect matched original/successor workloads with repeated client-frame and server-tick measurements. **No FPS improvement is demonstrated**; the [comparison tool](performance-comparison.md) has synthetic regression evidence only. Complete provenance/dependency review, obtain a clean native build and satisfy the [release checker](release-qualification.md). The [packed Core startup](packed-core-startup.md) is now demonstrated for its exact archived PAK, while the [packaging shutdown/control failures](native-packaging-control.md) remain unresolved and unqualified. The [development source bundle](development-source-bundle.md) provides editable projects; it is separate from that native candidate and from a qualified release. No release package is published.

[Private test cleanup](evidence/private-prefix-cleanup-2026-09-19-1631/review.json) now verifies Proton’s trailing-slash prefix spelling, checks idle state under the shared lock, and reaps only positively identified adopted test-device helpers after owned shutdown. The latest companion and ordinary runs record verified cleanup. Earlier strict-byte idle checks could miss those helpers; their historical logs are retained unchanged.

Original missions remain untouched. Engine tests use private profiles/prefixes, hidden displays and muted test audio. RHS/WCS/BMP are now authorized dependencies for AI/full integration. Additional terrains, complete original feature parity, alternate starts and distant-unit materialization remain unfinished. The [validation ledger](validation.md) preserves the detailed historical checkpoints and failures.