diff --git a/PLANS/2026-01-06-messagepack-cache-conversion-design.md b/PLANS/2026-01-06-messagepack-cache-conversion-design.md
new file mode 100644
index 0000000..e51f04d
--- /dev/null
+++ b/PLANS/2026-01-06-messagepack-cache-conversion-design.md
@@ -0,0 +1,142 @@
+# MessagePack Cache Conversion Design
+
+## Purpose
+
+Convert the development cache files in `CACHED_DB_FILES/` from zstd-compressed JSON (`.json.zstd`) to zstd-compressed MessagePack (`.msgpack.zstd`) for faster deserialization and smaller file sizes.
+
+## Goals
+
+1. **Faster deserialization** - MessagePack is faster to parse than JSON
+2. **Smaller file sizes** - MessagePack is more compact than JSON
+
+## Current State
+
+- 22 cache files in `CACHED_DB_FILES/` totaling ~3.6 GB (zstd-compressed JSON)
+- `JsonZstdFileSource` reads files using ZstdSharp + Utf8JsonReader
+- Each `*DevEtl.cs` defines a schema and creates a pipeline from JSON files
+- Tests verify ETL loads data from cache files into SQL Server
+
+## Design Decisions
+
+| Decision | Choice | Rationale |
+|----------|--------|-----------|
+| Conversion approach | One-time manual | Cache files are static snapshots, not actively regenerated |
+| Data structure | Map format (field names as keys) | Self-describing, maintainable, keys compress well with zstd |
+| Compression | Keep zstd | Largest file is 878 MB; raw MessagePack would be 2-4x larger |
+| Converter location | Standalone console app in `Tools/CacheConverter/` | Isolated utility, not part of main solution |
+
+## File Format
+
+**New extension:** `.msgpack.zstd`
+
+**File naming:**
+- `branch.json.zstd` → `branch.msgpack.zstd`
+- `workordertime_curr.json.zstd` → `workordertime_curr.msgpack.zstd`
+
+**Data structure:** Array of maps (same logical structure as JSON)
+```
+[
+  { "Code": "ABC", "Description": "Branch ABC", "LastUpdateDT": <DateTime> },
+  { "Code": "DEF", "Description": "Branch DEF", "LastUpdateDT": <DateTime> },
+  ...
+]
+```
+
+**Library:** MessagePack-CSharp (`MessagePack` NuGet package)
+
+## Components
+
+### 1. Converter Tool
+
+**Location:** `/JdeScopingTool/Tools/CacheConverter/`
+
+```
+Tools/
+└── CacheConverter/
+    ├── CacheConverter.csproj
+    └── Program.cs
+```
+
+**Dependencies:**
+- `ZstdSharp.Port` - read zstd JSON, write zstd MessagePack
+- `MessagePack` - MessagePack serialization
+
+**Behavior:**
+1. Read each `.json.zstd` file from `CACHED_DB_FILES/`
+2. Stream JSON → deserialize to `Dictionary<string, object?>[]`
+3. Serialize to MessagePack (map format) → compress with zstd
+4. Write to `.msgpack.zstd` alongside originals
+5. Print before/after sizes for comparison
+
+**Usage:**
+```bash
+cd Tools/CacheConverter
+dotnet run -- ../../CACHED_DB_FILES
+```
+
+### 2. MessagePackZstdFileSource
+
+**New file:** `NEW/src/JdeScoping.DataSync.Dev/Sources/MessagePackZstdFileSource.cs`
+
+- Implements `IImportSource` (same interface as `JsonZstdFileSource`)
+- Reads `.msgpack.zstd` files using streaming decompression
+- Uses `MessagePackStreamReader` for efficient streaming deserialization
+- Returns an `IDataReader` that yields rows one at a time
+- Schema still needed for `IDataReader` field metadata (column names, types, ordinals)
+
+**Package addition:** Add `MessagePack` to `JdeScoping.DataSync.Dev.csproj`
+
+### 3. DevEtl Class Updates
+
+**Changes to each `*DevEtl.cs` file (22 files):**
+
+1. Update `CacheFileName` constant:
+   ```csharp
+   // Before
+   public static readonly string CacheFileName = "branch.json.zstd";
+   // After
+   public static readonly string CacheFileName = "branch.msgpack.zstd";
+   ```
+
+2. Update `Create()` method:
+   ```csharp
+   // Before
+   .WithSource(new JsonZstdFileSource(cacheFilePath, Schema))
+   // After
+   .WithSource(new MessagePackZstdFileSource(cacheFilePath, Schema))
+   ```
+
+**No changes to:**
+- Schema definitions (same column names and types)
+- Pipeline structure
+- `DevEtlRegistry.cs`
+
+### 4. Cleanup (After Verification)
+
+Remove obsolete JSON readers:
+- `JsonZstdFileSource.cs`
+- `JsonStreamingDataReader.cs`
+- `Utf8JsonStreamingDataReader.cs`
+
+Remove old cache files:
+- All `*.json.zstd` files in `CACHED_DB_FILES/`
+
+## Test Strategy
+
+1. Run converter tool, verify all 22 files convert without errors
+2. Compare file sizes (expect 10-30% reduction)
+3. Run existing `JdeScoping.DataSync.Dev.Tests` - all tests should pass unchanged
+4. Verify data loaded matches previous JSON-based loads
+
+## Files Changed
+
+| File | Change |
+|------|--------|
+| `Tools/CacheConverter/` (new) | Standalone converter tool |
+| `Sources/MessagePackZstdFileSource.cs` (new) | New MessagePack reader |
+| `JdeScoping.DataSync.Dev.csproj` | Add MessagePack package |
+| `*DevEtl.cs` (22 files) | Update file extension and source class |
+| `Sources/JsonZstdFileSource.cs` | Delete after migration |
+| `Sources/JsonStreamingDataReader.cs` | Delete after migration |
+| `Sources/Utf8JsonStreamingDataReader.cs` | Delete after migration |
+| `CACHED_DB_FILES/*.json.zstd` | Delete after verification |