Docs: document the new features, Discord screenshot placeholders, fix start scripts #3

Merged

Khyretos merged 2 commits from claude/practical-darwin-vrjmvc into main

2026-09-24 23:59:02 +02:00

Khyretos commented

2026-09-24 23:58:53 +02:00

(Migrated from github.com)

Summary

Documentation follow-up to #2, plus a fix for the start scripts.

README: documents the Discord voice channel source (features, audio sources, usage, Docker notes, compose snippet, project structure) and the other features from #2 (live Whisper mode, auto-stop on lost audio, self-healing browser audio, vertical position, saved popout ID, hallucination guard, native-rate input devices). Adds two screenshot placeholders, images/discord-settings.png and images/discord-popout.png, to be filled in later. Corrects the statement that nothing ever stops a running session, and the download_vosk_models.py examples.
Start scripts: start.bat, and the final message of setup.sh / setup.bat, ran python app.py, which doesn't exist; they now run voice_translator.py. Same fix in CONFIG_EXAMPLES.txt.
QUICKSTART: correct start command and Vosk download syntax, current audio sources, named sessions, Discord.
TROUBLESHOOTING: new sections for auto-stop, reconnects, a stale or hand-merged page script, WebSocket 404, missing modules, sample-rate errors, popout "Not found", Whisper lag and hallucinations, and Discord setup errors.
AUDIO_SOURCES / REFACTOR: Discord as the fourth source, auto-stop, the Dockerfile copying all modules, and the modules added since.

The test suite passes (157 tests); no code changes other than the start scripts.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Hr7di2bGC8qCojX7XXQbGx

Generated by Claude Code

## Summary Documentation follow-up to #2, plus a fix for the start scripts. - **README**: documents the Discord voice channel source (features, audio sources, usage, Docker notes, compose snippet, project structure) and the other features from #2 (live Whisper mode, auto-stop on lost audio, self-healing browser audio, vertical position, saved popout ID, hallucination guard, native-rate input devices). Adds two screenshot placeholders, `images/discord-settings.png` and `images/discord-popout.png`, to be filled in later. Corrects the statement that nothing ever stops a running session, and the `download_vosk_models.py` examples. - **Start scripts**: `start.bat`, and the final message of `setup.sh` / `setup.bat`, ran `python app.py`, which doesn't exist; they now run `voice_translator.py`. Same fix in `CONFIG_EXAMPLES.txt`. - **QUICKSTART**: correct start command and Vosk download syntax, current audio sources, named sessions, Discord. - **TROUBLESHOOTING**: new sections for auto-stop, reconnects, a stale or hand-merged page script, WebSocket 404, missing modules, sample-rate errors, popout "Not found", Whisper lag and hallucinations, and Discord setup errors. - **AUDIO_SOURCES / REFACTOR**: Discord as the fourth source, auto-stop, the Dockerfile copying all modules, and the modules added since. The test suite passes (157 tests); no code changes other than the start scripts. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01Hr7di2bGC8qCojX7XXQbGx --- _Generated by [Claude Code](https://claude.ai/code/session_01Hr7di2bGC8qCojX7XXQbGx)_