danvics/pediatric-ai-scribe-v3
1
0
Fork 0
Code Issues Pull requests Projects Releases 16 Packages Wiki Activity Actions

fix: /api/health/detailed reported a text-to-speech provider that does not exist

Browse source 
The endpoint answered from environment variables of its own rather than from the
speech code, so ELEVENLABS_API_KEY being set made it report tts: 'elevenlabs'.
There is no ElevenLabs path in ttsProvider.js — getTTSProvider() only ever
returns 'litellm' or 'none' — so the one endpoint an operator checks to find out
what speech is doing was reporting a provider this app cannot use. Likewise
whisper: OPENAI_API_KEY, which stopped describing STT when transcription moved
behind the gateway.

Both now ask getSTTProvider() and getTTSProvider(), the same functions the routes
ask, so the answer cannot drift from behaviour again.

Removed the two other ElevenLabs leftovers: the FAQ told users their notes were
read aloud by "Google, OpenAI, or ElevenLabs", and a TTS test used 'elevenlabs'
as its sample value for the passthrough of TTS_PROVIDER, which implied the
provider was supported.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Dv6sqaY6Vq3ChZHMem3cnU
This commit is contained in:
Daniel 2026-09-12 05:05:39 +02:00
parent 67e416c6d9
commit cfb8aab77d
3 changed files with 9 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
public/components/faq.html
View file

@ -124,7 +124,7 @@
<div class="faq-item">
<button class="faq-question">Can the AI read my notes aloud?</button>
<div class="faq-answer">
<p>Yes. Click the <strong>Read</strong> button on any generated note to hear it spoken aloud. This uses text-to-speech (TTS) powered by Google, OpenAI, or ElevenLabs depending on your setup. You can choose your preferred voice in <strong>Settings &gt; Voice Preferences</strong>.</p>
<p>Yes. Click the <strong>Read</strong> button on any generated note to hear it spoken aloud. Which voice engine does the speaking depends on what your administrator has configured. You can choose your preferred voice in <strong>Settings &gt; Voice Preferences</strong>.</p>
</div>
</div>
</div>

8
server.js
View file

@ -308,8 +308,12 @@ app.get('/api/health/detailed', _hcAuth, _hcAdmin, (req, res) => {
bedrock: process.env.AWS_BEDROCK_REGION ? 'configured' : 'not configured',
azure: process.env.AZURE_OPENAI_ENDPOINT ? 'configured' : 'not configured',
litellm: process.env.LITELLM_API_BASE ? 'configured' : 'not configured',
whisper: process.env.OPENAI_API_KEY ? 'configured' : 'missing',
tts: process.env.LITELLM_API_BASE ? 'litellm' : (process.env.ELEVENLABS_API_KEY ? 'elevenlabs' : 'none')
// Asked of the same functions the routes ask, so this cannot drift from what
// speech actually does. It used to answer from environment variables of its
// own and would report tts: 'elevenlabs' for a provider that does not exist
// in this app — a health check that lies is worse than one that is absent.
stt: require('./src/utils/sttProvider').getSTTProvider(),
tts: require('./src/utils/ttsProvider').getTTSProvider()
});
});

4
test/tts-provider.test.js
View file

@ -27,8 +27,8 @@ test('TTS provider defaults to LiteLLM when gateway is configured', () => {
test('TTS provider ignores non-LiteLLM provider overrides', () => {
const ttsProvider = require('../src/utils/ttsProvider');
withEnv({ TTS_PROVIDER: 'elevenlabs', LITELLM_API_BASE: 'https://llm.example.com' }, () => {
assert.equal(ttsProvider.getTTSEnvProvider(), 'elevenlabs');
withEnv({ TTS_PROVIDER: 'some-other-provider', LITELLM_API_BASE: 'https://llm.example.com' }, () => {
assert.equal(ttsProvider.getTTSEnvProvider(), 'some-other-provider');
assert.equal(ttsProvider.getTTSProvider(), 'litellm');
});
});