diff --git a/JAVA_MIGRATION.md b/JAVA_MIGRATION.md
index 5ccc83a..62e6b89 100644
--- a/JAVA_MIGRATION.md
+++ b/JAVA_MIGRATION.md
@@ -23,15 +23,15 @@ No trustworthy tool will produce a maintainable Java port automatically. A decom
 
 ## Current status
 
-The binary and source-level consumer audits are complete. Phase 0 still lacks exact description-packet fixtures and a
-pre-optimization profile. The low-coupling Phase 3 queue is complete; before entering measured hot-path or generated
-tile work, the immediate milestone is the audit-derived compatibility gate below.
+The binary and source-level consumer audits are complete. Phase 0 still lacks a pre-optimization profile. The
+low-coupling Phase 3 queue is complete; before entering measured hot-path work, the immediate milestone is the
+remaining generated-tile portion of the audit-derived compatibility gate below.
 
 That audit found and this branch has now repaired one existing regression: Schematica 1.12.6 reflects the original private
 `MultiPartRegistry$` field `codechicken$multipart$MultiPartRegistry$$typeMap` and casts it to a Scala mutable map. The
 companion again exposes that exact private field as a live view of the canonical Java `HashMap`, and a test performs
-Schematica's reflective lookup. The remaining source-only member guards and `TileMultipart` ordering/move lifecycle
-are also frozen; compact mixed-part NBT and description-packet fixtures are next.
+Schematica's reflective lookup. The remaining source-only member guards, `TileMultipart` ordering/move lifecycle, and
+compact mixed-part NBT/description packet are also frozen; generated-trait and pass-through fixtures are next.
 
 Phase 1 has not properly started: GTNHLib is present but was added at `api` scope to satisfy a test compile rather
 than because a migration change needed it, and no Forge mod dependency was declared. Revisit that entry only when a
@@ -305,8 +305,8 @@ Do this before another source conversion:
   `registerMaterial` lookup.
 - [x] Freeze `TileMultipart` order and lifecycle observations: `parts`/`id` NBT, `partList`/`jPartList`, `partMap`, slot
   rebinding, add/remove/replace callback order, moving the live tile, and `onMoved`.
-- [ ] Add compact mixed-part NBT and description-packet fixtures. Use representative fake parts in the existing test
-  harness; do not add all downstream mods as test dependencies.
+- [x] Add compact mixed-part NBT and description-packet fixtures. The existing torch/button parts provide the needed
+  distinct IDs, order, slots and payload without adding downstream mods or a test-only part registry.
 - [ ] Freeze one generated Scala trait, one Java trait, and representative pass-through interfaces before changing
   built-in tile traits or either generator.
 
@@ -394,7 +394,7 @@ The work should move from low to high risk while keeping a buildable mixed-langu
 Extensive tests are both possible and recommended here. Their purpose is not to prove that the current Scala behavior is ideal; it is to make its observable behavior explicit before translation, so every later difference is intentional.
 
 The repository now has a Java-8-compatible JUnit Jupiter suite and a small Forge integration runner. The current
-baseline is 130 plain-JVM tests and 24 Forge server tests. Minecraft 1.7.10 does not provide the modern GameTest
+baseline is 130 plain-JVM tests and 25 Forge server tests. Minecraft 1.7.10 does not provide the modern GameTest
 framework, so game-dependent coverage continues through the existing narrow Forge harness rather than a second test
 framework.
 
@@ -620,4 +620,7 @@ generated-trait compatibility work. Scala-runtime removal remains a later projec
 - Reproduced MatterManipulator's live-tile relocation shape. Removing the tile first calls `onWorldSeparate` for each
   part in list order; after reinsertion, `onMoved` calls each part's `onMoved` and default `onWorldJoin`, again in list
   order. The original generated tile, part instances, order, slot map and tile references survive.
-- The expanded baseline is 130 plain-JVM tests and 24 Forge server tests, all passing.
+- Froze the complete logical chunk-description payload around the same torch/button pair: packet type, chunk
+  coordinates, tile count, nibble-relative X/Z, full Y, part count, sorted runtime part IDs, part order and both meta
+  bytes. CodeChickenLib's deferred Deflater transport step is deliberately outside this FMP fixture.
+- The expanded baseline is 130 plain-JVM tests and 25 Forge server tests, all passing.
diff --git a/JAVA_MIGRATION_CONSUMER_AUDIT.md b/JAVA_MIGRATION_CONSUMER_AUDIT.md
index 028a587..476cf8f 100644
--- a/JAVA_MIGRATION_CONSUMER_AUDIT.md
+++ b/JAVA_MIGRATION_CONSUMER_AUDIT.md
@@ -445,8 +445,9 @@ The Java port may add faster indexed storage, but all published views, serializa
 packet consumers must observe one consistent order.
 
 Current-branch verification: `MultipartHelperFunctionalTest` freezes ordered two-part NBT reconstruction and slot
-rebinding. `TileMultipartLifecycleFunctionalTest` freezes add/remove callbacks and the MatterManipulator-shaped live
-move, including ordered `onWorldSeparate` before ordered `onMoved`/`onWorldJoin`, without replacing the generated tile.
+rebinding plus the exact logical chunk-description framing and part payload. `TileMultipartLifecycleFunctionalTest`
+freezes add/remove callbacks and the MatterManipulator-shaped live move, including ordered `onWorldSeparate` before
+ordered `onMoved`/`onWorldJoin`, without replacing the generated tile.
 
 ### Material IDs and packets
 
@@ -495,7 +496,8 @@ behavior:
 3. Keep the completed `TileMultipart` list/map/slot and live-move characterization green before optimizing storage or
    iteration.
 4. Characterize pass-through interfaces and ProjectRed's external Scala trait before changing either generator.
-5. Add mixed NBT/packet fixtures before converting microblocks and Minecraft part implementations.
+5. Keep the completed compact mixed NBT/chunk-description fixture green; add subsystem-specific payload cases when
+   converting microblocks or Minecraft part implementations.
 6. Treat Scala runtime removal as downstream coordination work after the Java port, not as a requirement of it.
 
 With these gates, the source audit is actionable: the port can simplify internals freely where no listed observation
diff --git a/JAVA_MIGRATION_HANDOFF.md b/JAVA_MIGRATION_HANDOFF.md
index 7852819..d5457f3 100644
--- a/JAVA_MIGRATION_HANDOFF.md
+++ b/JAVA_MIGRATION_HANDOFF.md
@@ -57,7 +57,7 @@ awk -F'"' '/<testsuite /{t+=$4;f+=$8;e+=$10} END{print "tests="t" failures="f" e
 grep -o 'tests="[0-9]*" skipped="[0-9]*" failures="[0-9]*" errors="[0-9]*"' run/server/junit-out/TEST-*.xml
 ```
 
-Current baseline: **130 plain-JVM tests, 24 Forge server tests, all passing at their last completed runs.**
+Current baseline: **130 plain-JVM tests, 25 Forge server tests, all passing at their last completed runs.**
 
 ### ABI diff against the reference
 
@@ -204,8 +204,11 @@ Tile list/order/slot behavior and the live move lifecycle are now frozen by focu
 pin ordered `parts`/`id` NBT, published list views, slot rebinding, add/remove/replace behavior, and the relocation
 sequence (`onWorldSeparate`, then `onMoved`/`onWorldJoin`) while preserving the original generated tile.
 
-Next add compact mixed-part NBT and exact description-packet fixtures, followed by representative generated-trait and
-pass-through fixtures. The checklist is in the "Immediate compatibility gate" section of `JAVA_MIGRATION.md`.
+The same ordered torch/button pair now supplies a compact NBT fixture and an exact logical chunk-description fixture.
+The latter pins FMP's framing and payload before CodeChickenLib applies deferred transport compression.
+
+Next freeze representative generated Scala/Java traits and pass-through interfaces. The checklist is in the
+"Immediate compatibility gate" section of `JAVA_MIGRATION.md`.
 
 `IRedstonePart.scala` is misleadingly named and is **not** a marker-trait file. It holds six traits plus
 `RedstoneInteractions`, whose `MODULE$` is load-bearing. After the compatibility gate and an initial profile, port the
@@ -228,8 +231,8 @@ all of `multipart/asm/`. The ASM subsystem should be last; freeze generated-clas
   Every port since `TCuboidPart` has deferred something to it, and `BlockMultipart` leans on it hardest: breaking, selection boxes, collision, pick block, activation, particles and
   light are covered by nothing automated. Run it in a real client before this branch goes near a release.
 - No CPU or allocation profiles exist. Phase 4 has not started, so there is no baseline to compare against.
-- Tile `parts`/`id` ordering now has a live NBT round-trip fixture, but broad part payload and exact description-packet
-  fixtures are still missing.
+- Microblock-specific NBT and packet payloads still need characterization immediately before that subsystem changes;
+  the compact core tile/part fixture is complete.
 - `TileMultipart`'s `partList` still republishes an immutable Scala `Seq` on every mutation, and `operate` still
   allocates a `Function1` per call. Both are faithful to the reference and both are deliberately left for Phase 4,
   which needs profiling first.
diff --git a/src/functionalTest/java/codechicken/multipart/test/MultipartHelperFunctionalTest.java b/src/functionalTest/java/codechicken/multipart/test/MultipartHelperFunctionalTest.java
index 207a074..12068d2 100644
--- a/src/functionalTest/java/codechicken/multipart/test/MultipartHelperFunctionalTest.java
+++ b/src/functionalTest/java/codechicken/multipart/test/MultipartHelperFunctionalTest.java
@@ -1,5 +1,6 @@
 package codechicken.multipart.test;
 
+import static org.junit.jupiter.api.Assertions.assertArrayEquals;
 import static org.junit.jupiter.api.Assertions.assertDoesNotThrow;
 import static org.junit.jupiter.api.Assertions.assertEquals;
 import static org.junit.jupiter.api.Assertions.assertInstanceOf;
@@ -17,16 +18,20 @@ import net.minecraft.nbt.NBTTagList;
 import net.minecraft.server.MinecraftServer;
 import net.minecraft.tileentity.TileEntity;
 import net.minecraft.world.World;
+import net.minecraft.world.chunk.Chunk;
 
 import org.junit.jupiter.api.Test;
 
+import codechicken.lib.packet.PacketCustom;
 import codechicken.multipart.MultiPartRegistry;
 import codechicken.multipart.MultipartHelper;
 import codechicken.multipart.TMultiPart;
 import codechicken.multipart.TileMultipart;
+import codechicken.multipart.handler.MultipartSPH;
 import codechicken.multipart.handler.MultipartSaveLoad;
 import codechicken.multipart.minecraft.ButtonPart;
 import codechicken.multipart.minecraft.TorchPart;
+import io.netty.buffer.ByteBuf;
 import scala.collection.JavaConversions;
 
 /**
@@ -55,9 +60,7 @@ class MultipartHelperFunctionalTest {
     @Test
     void rebuildsAnOrderedSlottedTileFromItsOwnNbtAndPublishesTheLoadingWorld() {
         World world = world();
-        TMultiPart torch = new TorchPart(5);
-        TMultiPart button = new ButtonPart(1);
-        TileMultipart saved = MultipartHelper.createTileFromParts(Arrays.asList(torch, button));
+        TileMultipart saved = mixedTile();
 
         NBTTagCompound tag = new NBTTagCompound();
         saved.writeToNBT(tag);
@@ -81,6 +84,26 @@ class MultipartHelperFunctionalTest {
         assertSame(world, MultipartSaveLoad.loadingWorld());
     }
 
+    @Test
+    void writesTheExactMixedPartChunkDescriptionPacket() {
+        World world = world();
+        Chunk chunk = world.getChunkFromChunkCoords(1, 2);
+        TileMultipart tile = mixedTile();
+        tile.xCoord = 18;
+        tile.yCoord = 200;
+        tile.zCoord = 35;
+
+        PacketCustom packet = MultipartSPH.getDescPacket(chunk, Collections.<TileEntity>singletonList(tile).iterator());
+        ByteBuf buffer = packet.getByteBuf();
+        byte[] actual = new byte[buffer.writerIndex()];
+        buffer.getBytes(0, actual);
+
+        // type | chunk x/z | tile count | local x/y/z | part count | torch id/meta | button id/meta
+        assertArrayEquals(
+                new byte[] { 2, 0, 0, 0, 1, 0, 0, 0, 2, 0, 1, 2, 0, 0, 0, (byte) 200, 3, 2, 3, 5, 0, 1 },
+                actual);
+    }
+
     /** The id guard runs before anything else, so a foreign tag must not disturb the loading world. */
     @Test
     void leavesTheLoadingWorldAloneForATagItDoesNotOwn() {
@@ -125,6 +148,10 @@ class MultipartHelperFunctionalTest {
         return world;
     }
 
+    private static TileMultipart mixedTile() {
+        return MultipartHelper.createTileFromParts(Arrays.asList(new TorchPart(5), new ButtonPart(1)));
+    }
+
     /** Matches only its own tile class, so registering it cannot affect any real chunk load. */
     private static final class MarkerTile extends TileEntity {
     }