ADR-224: Implement intent stages (PhraseVariator, VocabComputer, DVC entry-points)

.claude/settings.json

@@ -1,20 +1,20 @@
 {
+  "enabledPlugins": {
+    "dotnet@dotnet-agent-skills": false,
+    "dotnet-diag@dotnet-agent-skills": true,
+    "dotnet-aspnet@dotnet-agent-skills": false,
+    "dotnet-template-engine@dotnet-agent-skills": false,
+    "dev-team@jodavis-agent-plugins": true
+  },
   "extraKnownMarketplaces": {
-    "dev-team-agents": {
+    "jodavis-agent-plugins": {
       "source": {
         "source": "github",
         "repo": "jodavis/agent-plugins",
-        "ref": "feature/ADR-246-cloud-dev-team"
+        "ref": "feature/ADR-269-agent-orchestration"
       }
     }
   },
-  "enabledPlugins": {
-    "dotnet@dotnet-agent-skills": false,
-    "dotnet-diag@dotnet-agent-skills": true,
-    "dotnet-aspnet@dotnet-agent-skills": false,
-    "dotnet-template-engine@dotnet-agent-skills": false,
-    "dev-team@dev-team-agents": true
-  },
   "mcpServers": {
     "microsoft-learn": {
       "type": "sse",

.github/workflows/build-and-test.yml

@@ -75,8 +75,9 @@ jobs:
       with:
         python-version: '3.12'
 
-    - name: Install pytest
-      run: pip install pytest
+    - name: Install Python dependencies
+      working-directory: ml
+      run: pip install -r requirements.txt
 
     - name: Run Python unit tests
       working-directory: ml

Directory.Packages.props

@@ -20,6 +20,7 @@
     <!-- Third-party Libraries -->
     <PackageVersion Include="I8Beef.TiVo" Version="1.0.0.14" />
     <PackageVersion Include="Microsoft.Playwright" Version="1.58.0" />
+    <PackageVersion Include="MessagePack.Annotations" Version="2.5.198" />
     <PackageVersion Include="StreamJsonRpc" Version="2.24.84" />
     <PackageVersion Include="System.Speech" Version="8.0.0" />
     <PackageVersion Include="System.Text.Json" Version="8.0.5" />

ml/_spec_OopPipeline.md

@@ -985,6 +985,23 @@ Implement `ml/pipeline/intent/` and the first two DVC entry-points.
 
 ---
 
+### [ADR-278](https://jodasoft.atlassian.net/browse/ADR-278) Task 3b: DVC wiring for intent stages
+
+Wire the first two intent stages in DVC as an early end-to-end proof that the OOP classes, manifest format, and pipeline conventions work correctly before the speech stages are implemented. This task initialises DVC inside `ml/`, configures the S3 remote, creates `dvc.yaml` covering the phoneme dictionary download and the two intent stages, writes the initial `params.yaml` with the pipeline keys those stages need, and verifies that `dvc repro` runs both stages successfully and `dvc push` uploads their outputs to S3.
+
+- [ ] `dvc init` run inside `ml/` (creates `ml/.dvc/`); `.dvc/` added to `ml/.gitignore`
+- [ ] S3 remote configured in `ml/.dvc/config`: `dvc remote add -d adr-ml-training-data s3://adr-ml-training-data/dvc` (matches repo-root `.dvc/config`)
+- [ ] `ml/params.yaml` created with all `pipeline.*` keys needed by the intent stages: `input_phrases_path`, `variations_per_phrase`, `subsample_rate`
+- [ ] `ml/dvc.yaml` created with three stages in dependency order:
+  - `download_phoneme_dictionary` — wraps `download_phoneme_dictionary.py`; output dir tracked by DVC
+  - `generate_phrases` — runs `intent_01_generate_phrases.py`; reads params `pipeline.input_phrases_path`, `pipeline.variations_per_phrase`, `pipeline.subsample_rate`; deps include `pipeline/` package directory and input phrases file; out: `data/intent_01_generate_phrases/`
+  - `compute_vocab` — runs `intent_02_compute_vocab.py`; deps include `data/intent_01_generate_phrases/` and the phoneme dictionary output dir; out: `data/intent_02_compute_vocab/`
+- [ ] `dvc repro` runs both intent stages end-to-end without errors on the dev machine
+- [ ] `dvc push` successfully uploads intent stage outputs to the S3 remote
+- [ ] `validate-build` and `validate-tests` pass
+
+---
+
 ### [ADR-225](https://jodasoft.atlassian.net/browse/ADR-225) Task 4: TTS stage
 
 Implement `ml/pipeline/speech/tts_stage.py` and `stages/speech_03_generate_samples.py`.
@@ -1061,12 +1078,12 @@ Implement `ModelEvaluator` and the final two entry-points.
 
 ### [ADR-231](https://jodasoft.atlassian.net/browse/ADR-231) Task 9: DVC wiring and old script cleanup
 
-Wire all stages in `dvc.yaml`; migrate `params.yaml`; delete old scripts.
+Extend the `dvc.yaml` and `params.yaml` from Task 3b to cover all 12 stages; delete old scripts.
 
 **⚠️ Script deletion must happen after Task 3 is merged** — the old `ml/scripts/` tree is the reference for `PhraseVariator`. Delete it only once Task 3's implementation is reviewed and merged.
 
-- [ ] Write `ml/dvc.yaml` from scratch: all 12 stages wired with correct `cmd`, `deps`, `outs`, `params`; `persist: true` on all `ModifierStage` output dirs
-- [ ] Write `ml/params.yaml`: all `pipeline.*` keys (including `input_phrases_path`, `variations_per_phrase`, `subsample_rate`, `n_mels`, `time_steps`, `input_token_length`, `epochs`, `batch_size`) and all `stages.*` variation-constraint keys
+- [ ] Extend `ml/dvc.yaml` with the remaining 9 stages (speech_03 through speech_12): correct `cmd`, `deps`, `outs`, `params`; `persist: true` on all `ModifierStage` output dirs
+- [ ] Extend `ml/params.yaml` with remaining `pipeline.*` keys (`n_mels`, `time_steps`, `input_token_length`, `epochs`, `batch_size`) and all `stages.*` variation-constraint keys
 - [ ] Retain `download_phoneme_dictionary.py` as a non-OOP stage
 - [ ] Delete `ml/scripts/` tree (after Task 3 is merged)
 - [ ] `dvc repro` runs end-to-end without errors on the dev machine

ml/params.yaml

@@ -0,0 +1,11 @@
+pipeline:
+  variations_per_phrase: 20
+  subsample_rate: 200
+
+stages:
+  generate_phrases:
+    repeat_modifier_chance: 0.2
+    pleasantry_chance: 0.3
+    hesitation_chance: 0.3
+    spelling_variant_chance: 0.1
+    case_variant_chance: 0.2