From 8b2bf136f17370561bb08a70112d856211c812c0 Mon Sep 17 00:00:00 2001
From: Jon Saad-Falcon <jonsaadfalcon@gmail.com>
Date: Tue, 3 Mar 2026 05:51:29 +0000
Subject: [PATCH] feat(speech): add SpeechConfig to configuration system

Add SpeechConfig dataclass with backend, model, language, device, and
compute_type fields. Wire it into JarvisConfig and load_config() so
[speech] TOML sections are loaded automatically.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---
 src/openjarvis/core/config.py | 14 ++++++++++++++
 tests/speech/__init__.py      |  0
 tests/speech/test_config.py   | 19 +++++++++++++++++++
 3 files changed, 33 insertions(+)
 create mode 100644 tests/speech/__init__.py
 create mode 100644 tests/speech/test_config.py

diff --git a/src/openjarvis/core/config.py b/src/openjarvis/core/config.py
index 1d16f373..b4a0419d 100644
--- a/src/openjarvis/core/config.py
+++ b/src/openjarvis/core/config.py
@@ -830,6 +830,17 @@ class OperatorsConfig:
     auto_activate: str = ""  # Comma-separated operator IDs
 
 
+@dataclass(slots=True)
+class SpeechConfig:
+    """Speech-to-text settings."""
+
+    backend: str = "auto"  # "auto", "faster-whisper", "whisper-cpp", "openai", "deepgram"
+    model: str = "base"  # Whisper model size: tiny, base, small, medium, large-v3
+    language: str = ""  # Empty = auto-detect
+    device: str = "auto"  # "auto", "cpu", "cuda"
+    compute_type: str = "float16"  # "float16", "int8", "float32"
+
+
 @dataclass
 class JarvisConfig:
     """Top-level configuration for OpenJarvis."""
@@ -851,6 +862,7 @@ class JarvisConfig:
     sessions: SessionConfig = field(default_factory=SessionConfig)
     a2a: A2AConfig = field(default_factory=A2AConfig)
     operators: OperatorsConfig = field(default_factory=OperatorsConfig)
+    speech: SpeechConfig = field(default_factory=SpeechConfig)
 
     @property
     def memory(self) -> StorageConfig:
@@ -945,6 +957,7 @@ def load_config(path: Optional[Path] = None) -> JarvisConfig:
             "server", "telemetry", "traces", "security",
             "channel", "tools", "sandbox", "scheduler",
             "workflow", "sessions", "a2a", "operators",
+            "speech",
         )
         for section_name in top_sections:
             if section_name in data:
@@ -1196,6 +1209,7 @@ __all__ = [
     "SessionConfig",
     "SignalChannelConfig",
     "SlackChannelConfig",
+    "SpeechConfig",
     "StorageConfig",
     "TeamsChannelConfig",
     "TelegramChannelConfig",
diff --git a/tests/speech/__init__.py b/tests/speech/__init__.py
new file mode 100644
index 00000000..e69de29b
diff --git a/tests/speech/test_config.py b/tests/speech/test_config.py
new file mode 100644
index 00000000..804af199
--- /dev/null
+++ b/tests/speech/test_config.py
@@ -0,0 +1,19 @@
+"""Tests for speech configuration."""
+
+from openjarvis.core.config import JarvisConfig, SpeechConfig
+
+
+def test_speech_config_defaults():
+    cfg = SpeechConfig()
+    assert cfg.backend == "auto"
+    assert cfg.model == "base"
+    assert cfg.language == ""
+    assert cfg.device == "auto"
+    assert cfg.compute_type == "float16"
+
+
+def test_jarvis_config_has_speech():
+    cfg = JarvisConfig()
+    assert hasattr(cfg, "speech")
+    assert isinstance(cfg.speech, SpeechConfig)
+    assert cfg.speech.backend == "auto"