From ed7bd05530877f9c812d6146d37e483842268bcf Mon Sep 17 00:00:00 2001
From: nesquena-hermes <nesquena-hermes@users.noreply.github.com>
Date: Tue, 16 Jun 2026 17:21:11 +0000
Subject: [PATCH] fix(voice): recognize command STT providers in capability
 probe (#4312)

The /api/transcribe/capability probe's resolve_provider() only knew the
built-in STT providers, so an explicit stt.provider mapping to a Hermes
command-type provider was reported unavailable and the browser fell back
unnecessarily. Add command_provider_available(), delegating to the agent's
_resolve_command_stt_provider_config(provider, cfg) as the single source of
truth, consulted as the final fallthrough after the built-in chain (the
agent rejects built-in names internally, so ordering can't shadow a native
provider). Added a clarifying comment that command providers are
intentionally omitted from the auto-detect tuple to match the agent's
_get_provider() precedence.

Co-authored-by: LLMinem <LLMinem@users.noreply.github.com>
---
 api/upload.py                           | 15 +++++++++++++
 tests/test_voice_transcribe_endpoint.py | 28 +++++++++++++++++++++++++
 2 files changed, 43 insertions(+)

diff --git a/api/upload.py b/api/upload.py
index 3125cb7e..c9239518 100644
--- a/api/upload.py
+++ b/api/upload.py
@@ -457,6 +457,13 @@ def _stt_provider_capability_from_module(stt):
             # non-WAV input through ffmpeg before invoking the command.
             return has_local_command() and has_browser_audio_converter()
 
+        def command_provider_available(provider):
+            resolver = getattr(stt, "_resolve_command_stt_provider_config", None)
+            try:
+                return callable(resolver) and resolver(provider, cfg_dict) is not None
+            except Exception:
+                return False
+
         def resolve_provider(provider):
             if provider == "local":
                 if bool(getattr(stt, "_HAS_FASTER_WHISPER", False)):
@@ -485,6 +492,8 @@ def _stt_provider_capability_from_module(stt):
                     return "none"
             if provider == "elevenlabs":
                 return "elevenlabs" if bool(env("ELEVENLABS_API_KEY")) else "none"
+            if command_provider_available(provider):
+                return provider
             return "none"
 
         explicit = "provider" in cfg_dict
@@ -494,6 +503,12 @@ def _stt_provider_capability_from_module(stt):
             return provider != "none", provider if provider != "none" else configured
 
         for candidate in ("local", "local_command", "groq", "openai", "mistral", "xai", "elevenlabs"):
+            # Command (custom) STT providers are intentionally omitted from this
+            # auto-detect tuple to mirror the agent's _get_provider() (transcription_tools.py),
+            # which only auto-selects local > groq > openai and never auto-picks a command
+            # provider. A command-backed STT activates only via an explicit stt.provider.
+            # Do NOT add command providers here without matching the agent, or the WebUI
+            # probe will diverge from what the agent actually resolves.
             provider = resolve_provider(candidate)
             if provider != "none":
                 return True, provider
diff --git a/tests/test_voice_transcribe_endpoint.py b/tests/test_voice_transcribe_endpoint.py
index 1374ebda..0126e94d 100644
--- a/tests/test_voice_transcribe_endpoint.py
+++ b/tests/test_voice_transcribe_endpoint.py
@@ -156,3 +156,31 @@ def test_handle_transcribe_capability_reports_actual_local_command_fallback(monk
 
     assert available is True
     assert provider == "local_command"
+
+
+def test_handle_transcribe_capability_reports_named_command_provider(monkeypatch):
+    stt_config = {
+        "provider": "elevenlabs-gemini",
+        "providers": {
+            "elevenlabs-gemini": {
+                "type": "command",
+                "command": "/home/user/.hermes/scripts/hermes-elevenlabs-stt.sh {input}",
+            }
+        },
+    }
+    fake_mod = types.ModuleType("tools.transcription_tools")
+    fake_mod.__dict__.update(
+        {
+            "_load_stt_config": lambda: stt_config,
+            "is_stt_enabled": lambda cfg=None: True,
+            "_HAS_FASTER_WHISPER": False,
+            "_HAS_OPENAI": False,
+            "_HAS_MISTRAL": False,
+            "_resolve_command_stt_provider_config": lambda provider, cfg: cfg["providers"].get(provider),
+        }
+    )
+
+    available, provider = _stt_provider_capability_from_module(fake_mod)
+
+    assert available is True
+    assert provider == "elevenlabs-gemini"