This commit brings several improvements to the application:
- Updates all Python dependencies in requirements.txt to their latest versions.
- Enhances file handling in capture.py by writing to a temporary file before renaming, preventing partial reads.
- Strengthens error handling for API calls (OpenAI, ElevenLabs) and file operations in both capture.py and narrator.py.
- Makes the ElevenLabs Voice ID configurable via an ELEVEN_VOICE_ID environment variable in narrator.py, with a sensible default.
- Aligns the narrator's persona in narrator.py with a "David Attenborough" style by updating the system prompt.
- Updates the README.md to remove outdated information, clarify API key usage, and include new configuration options.
- Confirms that the current audio saving mechanism is suitable for archival/logging.
- Upgrades the OpenAI model to gpt-4-turbo in narrator.py.
- Reduces console noise by making the "Say cheese!" message in capture.py print only once.
I did not add comprehensive docstrings and comments in this pass.