By name: speak description: Speak text aloud using the Magpie TTS container in the realtime-api Docker stack. Zero external dependencies — just curl + aplay. triggers: - speak - say - tell me - read aloud - announce
Speak via Magpie TTS
Synthesize speech using the local Magpie TTS container (port 9000) and play through speakers.
When to use
- When the user says "speak", "say", "tell me", "read aloud"
- After completing a significant task — speak a brief 1-2 sentence summary
- When the user asks you to announce or vocalize something
- To give audio feedback on deploy/test results
Usage
/home/spark/git/autonomous-intelligence/realtime-api/scripts/speak.sh "Text to speak"
Options
| Flag | Short | Default | Description |
|---|---|---|---|
| (positional) | (required) | Text to speak | |
--voice |
-v |
Mia.Calm |
Voice name |
--speed |
-s |
125 |
Speed percentage |
--url |
-u |
http://localhost:9000 |
Magpie TTS URL |
Voices
Mia, Mia.Calm, Mia.Happy, Mia.Sad, Mia.Angry, Aria, Aria.Calm, Aria.Happy, Jason, Jason.Calm, Leo, Leo.Calm
Examples
# Default calm voice
/home/spark/git/autonomous-intelligence/realtime-api/scripts/speak.sh "Deploy complete. All seven checks passed."
# Happy announcement
/home/spark/git/autonomous-intelligence/realtime-api/scripts/speak.sh -v Mia.Happy "All tests passed!"
# Different speed
/home/spark/git/autonomous-intelligence/realtime-api/scripts/speak.sh -s 140 "Speaking faster now."
Notes
- Uses the Magpie TTS container from the realtime-api docker-compose stack (localhost:9000)
- No Python dependencies — just curl and aplay
- Audio plays through PipeWire/ALSA
- Keep messages concise (1-2 sentences) for task summaries