From 0cc01c1450e29dbb03062771c56ca504d36aa153 Mon Sep 17 00:00:00 2001
From: Marius Mutu <mmarius28@gmail.com>
Date: Wed, 27 May 2026 14:44:13 +0000
Subject: [PATCH] =?UTF-8?q?feat(voice):=20Pas=2010=20=E2=80=94=20eco=20doc?=
 =?UTF-8?q?tor=20voice=20stack=20checks?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

cli.py: +101 / -0, append 9 checks după existing 15:
1. libopus loaded by discord.py (load_default fallback)
2. ffmpeg in PATH
3. Supertonic TTS reachable :7788 (5s timeout POST)
4. faster-whisper importable (no model load — too slow for doctor)
5. silero-vad importable
6. discord.ext.voice_recv importable (vendor package guard)
7-9. assets/voice/{thinking,beep_200ms,mhm}.wav exist + size thresholds

Helper _voice_doctor_checks() returns list[tuple[str, bool]] matching
doctor's reporting style. Replicates voice_setup.py logic in doctor
format (voice_setup uses ANSI colors directly, doctor uses (label, ok)
tuples — separate Option B implementation). Graceful ImportError
handling per check — never crashes the rest of eco doctor.

Exit code 1 corectly surfaces missing libopus (Discord voice silent
without it). Use `sudo apt install -y libopus0` to clear.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---
 cli.py | 101 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 101 insertions(+)

diff --git a/cli.py b/cli.py
index c5419c2..348e7c3 100755
--- a/cli.py
+++ b/cli.py
@@ -114,6 +114,104 @@ def _load_sessions_file() -> dict:
         return {}
 
 
+def _voice_doctor_checks() -> list[tuple[str, bool]]:
+    """Voice-stack health checks (Pas 10).
+
+    Mirrors the logic in tools/voice_setup.py but returns (label, ok) tuples
+    so they integrate with cmd_doctor's PASS/FAIL output. All checks degrade
+    gracefully — ImportError on optional voice deps is reported as FAIL, never
+    raised, so the rest of `eco doctor` is unaffected.
+    """
+    import importlib.util
+    import json as _json
+    import urllib.error
+    import urllib.request
+
+    results: list[tuple[str, bool]] = []
+
+    # 1. libopus0 loaded by discord.py
+    try:
+        import discord
+        if not discord.opus.is_loaded():
+            try:
+                discord.opus._load_default()
+            except Exception:
+                pass
+        results.append(("libopus loaded (discord.py)", discord.opus.is_loaded()))
+    except ImportError:
+        results.append(("libopus loaded (discord.py)", False))
+    except Exception:
+        results.append(("libopus loaded (discord.py)", False))
+
+    # 2. ffmpeg in PATH
+    results.append(("ffmpeg in PATH", shutil.which("ffmpeg") is not None))
+
+    # 3. Supertonic TTS reachable at http://127.0.0.1:7788/
+    supertonic_url = "http://127.0.0.1:7788/v1/audio/speech"
+    supertonic_ok = False
+    try:
+        payload = _json.dumps({
+            "model": "supertonic-3",
+            "input": "test",
+            "voice": "M2",
+            "response_format": "wav",
+            "lang": "ro",
+        }).encode("utf-8")
+        req = urllib.request.Request(
+            supertonic_url,
+            data=payload,
+            headers={"Content-Type": "application/json"},
+            method="POST",
+        )
+        with urllib.request.urlopen(req, timeout=5) as resp:
+            supertonic_ok = resp.status == 200
+    except (urllib.error.URLError, ConnectionError, OSError):
+        supertonic_ok = False
+    except Exception:
+        supertonic_ok = False
+    results.append(("Supertonic TTS reachable at :7788", supertonic_ok))
+
+    # 4. faster-whisper importable (don't load model — too slow)
+    results.append((
+        "faster-whisper importable",
+        importlib.util.find_spec("faster_whisper") is not None,
+    ))
+
+    # 5. silero-vad importable
+    results.append((
+        "silero-vad importable",
+        importlib.util.find_spec("silero_vad") is not None,
+    ))
+
+    # 6. discord.ext.voice_recv importable (vendor package)
+    voice_recv_ok = False
+    try:
+        voice_recv_ok = importlib.util.find_spec("discord.ext.voice_recv") is not None
+    except (ImportError, ValueError, ModuleNotFoundError):
+        voice_recv_ok = False
+    except Exception:
+        voice_recv_ok = False
+    results.append(("discord.ext.voice_recv importable", voice_recv_ok))
+
+    # 7-9. Voice assets present and non-trivial size
+    voice_assets = [
+        ("assets/voice/thinking.wav", 1024),
+        ("assets/voice/beep_200ms.wav", 512),
+        ("assets/voice/mhm.wav", 512),
+    ]
+    for rel_path, min_bytes in voice_assets:
+        path = PROJECT_ROOT / rel_path
+        ok = False
+        try:
+            ok = path.exists() and path.stat().st_size > min_bytes
+        except OSError:
+            ok = False
+        label = f"{rel_path} (>{min_bytes}B)"
+        results.append((label, ok))
+
+    return results
+
+
 def cmd_doctor(args):
     """Run diagnostic checks."""
     import re
@@ -227,6 +325,9 @@ def cmd_doctor(args):
     else:
         checks.append(("WhatsApp bridge (optional)", True))
 
+    # ---- Voice stack checks (Pas 10) ----
+    checks.extend(_voice_doctor_checks())
+
     # Print results
     all_pass = True
     for label, passed in checks: