How xenometon/ultron-ai works
Ultron is a small, single-file voice assistant. Its runtime starts with a spoken greeting, repeatedly captures and recognizes speech, then dispatches the recognized text to Wikipedia lookup, website opening, time reporting, or conversational responses; spoken output is produced through text-to-speech. The README also advertises general information and customizable commands, but the shown code has no separate information-search subsystem. All implementation nodes point to Codex.py; the graph reflects only the excerpted source and README.
Main entry points: Codex.py Command dispatch
Components 14
Other components
- Voice user
- Assistant loop
Codex.py - Greeting
Codex.py - Voice capture
Codex.py - Speech recognition
- Command dispatch
Codex.py - Wikipedia lookup
Codex.py - Time reporting
Codex.py - Assistant responses
Codex.py - Speech output
Codex.py - Wikipedia service
- Web browser
- System clock
- Requested websites
Connections 18
- Voice user → Voice capture: speaks commands
Codex.py - Assistant loop → Greeting: calls
Codex.py - Assistant loop → Voice capture: calls
Codex.py - Voice capture → Speech recognition: recognizes audio
Codex.py - Voice capture → Command dispatch: returns query
Codex.py - Assistant loop → Command dispatch: routes query
Codex.py - Command dispatch → Wikipedia lookup: calls
Codex.py - Command dispatch → Time reporting: reports time
Codex.py - Command dispatch → Assistant responses: handles phrases
Codex.py - Command dispatch → Web browser: opens pages
Codex.py - Wikipedia lookup → Wikipedia service: requests summary
Codex.py - Web browser → Requested websites: opens URL
- Greeting → System clock: reads hour
Codex.py - Time reporting → System clock: reads current time
Codex.py - Greeting → Speech output: speaks greeting
Codex.py - Wikipedia lookup → Speech output: speaks result
Codex.py - Time reporting → Speech output: speaks time
Codex.py - Assistant responses → Speech output: speaks reply
Codex.py