docs: #28 aurora connected gate USER-PASSED (intensity delta accepted); file #428 sky-slot pose-leak diagnostic

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

docs/ISSUES.md

@@ -24,6 +24,38 @@ What does NOT go here:
 - Every session: scan OPEN issues at start; promote/close anything we touched during the session before ending.
 - Promoting to a Phase: mark as `DONE (promoted to Phase X)` + commit SHA where the Phase entry landed.
 
+## #428 — Sky lightning slot fired 3 CreateParticle hooks with no live pose after a TimeSync day-group flip
+
+**Status:** OPEN
+**Severity:** LOW (bounded diagnostic spam; no crash, no leak)
+**Filed:** 2026-08-23 (launch11.err, first production run of the sky default-script port)
+**Component:** sky / vfx / script lifecycle
+
+**Symptom:** three identical stderr lines during the startup TimeSync window:
+`vfx: No live effect pose for owner 0xF8000007, part 0; emitter 0x320002C2 was not created.`
+Owner `0xF8000007` = post-scene sky slot index 7 — the Rainy lightning carrier
+(`0x02000BA6`, PES `0x33000453`, whose t=0 hooks are a `CallPES` self-loop
+(pause 30) plus `CreateParticle 0x320002C2 part=0`). The client boots on the
+pre-sync default calendar (PY10 day0, group 16 Rainy); the first TimeSync flips
+to the server date and the lightning slot vanishes, and hook executions landed
+AFTER `SkyPesFrameController` had already removed the slot's pose.
+
+**Analysis so far:** `PhysicsScriptRunner.StopAllForEntity` purges owners,
+anchors AND `_delayedCalls`, so this cannot loop forever (observed exactly 3,
+consistent with hooks already drained into the sink's queue, or an ordering
+window between the controller's stop and the runner's tick inside the same
+frame). Worth pinning: whether hook dispatch can outlive the owner's stop by
+one tick, and whether `part=0` hooks against a root-only synthetic pose (empty
+part list) resolve to the root as retail would. Repro: launch while the server
+clock is far from the pre-sync default during a Rainy-window fraction; watch
+stderr through the first TimeSync.
+
+**Files:** `src/AcDream.App/Rendering/SkyPesFrameController.cs`;
+`src/AcDream.Core/Vfx/PhysicsScriptRunner.cs` (`StopAllForEntity`, tick order);
+`src/AcDream.Core/Vfx/ParticleHookSink.cs` (`TryResolveAnchor`, part fallback).
+
+---
+
 ## #427 — Hard line below the horizon from altitude: the sky dome was fogged with a 0.2 floor and the world fog range came from the streaming window, not the keyframe
 
 **Status:** ✅ FIXED 2026-08-23 (owner report, Candeth Keep/Holtburg heights: "a cut off where the sky ends and the world background void begins").
@@ -16142,7 +16174,13 @@ additional flash-shader work).
 
 ## #28 — Aurora ("northern lights") effect not rendered
 
-**Status:** FIX LANDED 2026-08-23 — awaiting the connected visual gate
+**Status:** DONE — connected gate USER-PASSED 2026-08-23 (commit `18fce7bb`).
+The accepted residual: the aurora reads somewhat more intense in acdream than
+retail — the emitters are verbatim; the delta is the linear-light compositing
+(+ optional Atmospheric-pack bloom) brightening faint additive glows over a
+dark sky vs retail's gamma-space add. Accepted by the user ("gate pass for
+now. Might reopen it!") — if reopened, the principled fix is a gamma-faithful
+compensation on ADDITIVE particle contributions, never an aurora-only factor.
 **Severity:** LOW (aesthetic feature-parity)
 **Filed:** 2026-04-26
 **Component:** sky / vfx