Commands

Run each as python -m audioanalyser <flag>; one task per run.

Flag	What it does	Example
`-rec [settings.json]`	Record from the microphone	`-rec`
`-stt <file>`	Transcribe an audio file with Azure AI Speech	`-stt meeting.wav`
`-ta`	Analyse every transcript in `TRANSCRIPTS_FOLDER`	`-ta`
`-sum`	Summarise transcripts and write recommendations	`-sum`
`-t <codes...>`	Translate transcripts into these languages	`-t fr de`
`-tts <text> <name>`	Speak the text and save it as `<name>`	`-tts "hello" greeting`
`-s`	Start the dashboard on port 8080	`-s`
`-h`	Print the usage	`-h`

Configuration

Settings are environment variables, read from a .env file in the working directory. env.example documents most of them.

Variable	Used by
`INPUT_FOLDER`	Recording (output), transcription (input)
`TRANSCRIPTS_FOLDER`	Transcription (output), analysis, recommendations, translation (input)
`REPORTS_FOLDER`	Analysis (output)
`RECOMMENDATIONS_FOLDER`	Recommendations (output)
`TRANSLATIONS_FOLDER`	Translation (output)
`RECORDS_FOLDER`	Text to speech (output)

Variable	Service
`AZURE_AUDIO_TEXT_KEY`, `REGION`, `AUDIO_EXTENSION`	Azure AI Speech
`AZURE_LANGUAGE_ENDPOINT`, `AZURE_LANGUAGE_KEY`	Azure AI Language
`AZURE_TRANSLATOR_KEY`, `AZURE_TRANSLATOR_ENDPOINT`, `TRANSLATIONS_LANGUAGES`	Azure AI Translator
`TRANSCRIPTS_DB_TABLE_NAME`, `ANALYSIS_DB_TABLE_NAME`	SQLite table names

Variable	Meaning
`LLM_PROVIDER`	`openai` (default), `anthropic`, `gemini`, `ollama`, `claude-cli`, `agy-cli`
`LLM_MODEL`	Model for the chosen provider; overrides its default
`GPT3_API_KEY`, `OPENAI_MODEL`	OpenAI key, and the older model setting
`ANTHROPIC_API_KEY`	Anthropic key
`GEMINI_API_KEY`, `GOOGLE_CLOUD_PROJECT`, `GOOGLE_CLOUD_LOCATION`	Gemini, by key or through Vertex AI
`OLLAMA_HOST`	Ollama server, `http://localhost:11434` by default

CHANNELS, CHUNK, FORMAT, RATE and RECORD_SECONDS set the recording defaults; a settings file passed to -rec overrides them.