From e54df5f4f79ddda675a828ce1b0f06f1a7eeeee1 Mon Sep 17 00:00:00 2001
From: Vuong Hoang <vh@phasefinal.com>
Date: Mon, 27 Apr 2026 16:04:39 -0700
Subject: [PATCH] chatterbox: switch health probe from /health (doesn't exist)
 to /api/model-info
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

devnen/Chatterbox-TTS-Server doesn't expose /health — neither in code
nor OpenAPI. The deploy hung on the playbook's `Wait for /health to
respond` loop indefinitely (each curl -> 404, retry forever) even
though the container was up and the model loaded clean to CUDA at
22:52:21 (~42s after start).

/api/model-info returns `{"loaded":true,...}` only after the model
finishes loading, so it doubles as liveness + readiness. Updated:

* compose.yaml healthcheck — grep for `"loaded":true` from
  /api/model-info.
* playbook wait step — same probe instead of /health.
* verify /health → verify /api/model-info reports loaded.
* verify /v1/audio/voices — switched from greping for `voice|alloy|echo`
  literals to parsing JSON and asserting the actual response shape:
  `{"status":"ok","voices":[...]}` (devnen's shape — note this is NOT
  the OpenAI list-format vibevoice uses).
---
 playbooks/deploy-chatterbox.yaml | 22 ++++++++++++++++------
 stacks/chatterbox/compose.yaml   |  7 +++++--
 2 files changed, 21 insertions(+), 8 deletions(-)

diff --git a/playbooks/deploy-chatterbox.yaml b/playbooks/deploy-chatterbox.yaml
index d3f2d51..4bbbe72 100644
--- a/playbooks/deploy-chatterbox.yaml
+++ b/playbooks/deploy-chatterbox.yaml
@@ -70,22 +70,32 @@ steps:
   - name: docker compose up -d
     shell: cd {{ compose_dir }} && docker compose up -d
 
-  - name: Wait for /health to respond (allow ~15 min for model download + warmup)
+  - name: Wait for /api/model-info to report loaded:true (allow ~15 min for model download + warmup)
+    # The devnen wrapper has no /health endpoint. /api/model-info returns
+    # `{"loaded":true,...}` only after the model finishes loading, so it
+    # doubles as a liveness + ready probe.
     shell: |
       for i in $(seq 1 180); do
-        curl -sf -o /dev/null --max-time 3 http://localhost:{{ host_port }}/health && exit 0
+        curl -sf --max-time 3 http://localhost:{{ host_port }}/api/model-info \
+          | grep -q '"loaded":true' && exit 0
         sleep 5
       done
       exit 1
     changed_when: "false"
 
 verify:
-  - name: /health returns 200
-    shell: curl -sf -o /dev/null http://localhost:{{ host_port }}/health
+  - name: /api/model-info reports loaded:true
+    shell: curl -sf http://localhost:{{ host_port }}/api/model-info | grep -q '"loaded":true'
     changed_when: "false"
 
-  - name: /v1/audio/voices returns valid JSON
-    shell: curl -sf http://localhost:{{ host_port }}/v1/audio/voices | grep -q 'voice\|alloy\|echo'
+  - name: /v1/audio/voices returns the wrapper's status:ok + voices list
+    # Devnen's response shape is `{"status":"ok","voices":["Abigail.wav",...]}`
+    # — voices is an array of files in the predefined-voices dir. Assert
+    # both fields rather than greping for any one voice name (the seed
+    # set could change with upstream).
+    shell: |
+      curl -sf http://localhost:{{ host_port }}/v1/audio/voices \
+        | python3 -c "import json,sys; d=json.load(sys.stdin); assert d.get('status')=='ok' and isinstance(d.get('voices'), list)"
     changed_when: "false"
 
   - name: Container is running
diff --git a/stacks/chatterbox/compose.yaml b/stacks/chatterbox/compose.yaml
index 3527eb3..be101f4 100644
--- a/stacks/chatterbox/compose.yaml
+++ b/stacks/chatterbox/compose.yaml
@@ -49,8 +49,11 @@ services:
       # is to use the in-image config + env var overrides.
       # - ${CHATTERBOX_CONFIG}:/app/config.yaml:ro
     healthcheck:
-      # Devnen wrapper exposes /health; the OpenAPI/docs path is /docs.
-      test: ["CMD-SHELL", "curl -fsS -o /dev/null http://localhost:8004/health || exit 1"]
+      # The devnen wrapper does NOT expose /health (no such route — use
+      # /docs or /openapi.json to enumerate). /api/model-info returns
+      # `{"loaded":true,...}` only after the model finishes loading,
+      # so it doubles as a liveness + ready probe.
+      test: ["CMD-SHELL", "curl -fsS http://localhost:8004/api/model-info | grep -q '\"loaded\":true' || exit 1"]
       interval: 30s
       timeout: 10s
       retries: 3