gstack has eyes (
/browse) but no ears. This adds/listen— real-time system audio capture via PipeWire/PulseAudio for Linux environments.
A gstack skill for capturing system audio on Linux.
python3 capture.py # 5 seconds -> captured.wav
python3 capture.py 10 # 10 seconds -> captured.wav
python3 capture.py 5 out.wav
capture.py auto-detects the audio backend:
| Environment | Backend | Tool | Status |
|---|---|---|---|
| Ubuntu 24.04+ | PipeWire | pw-record --target |
Verified |
| Older Ubuntu / Debian | PulseAudio | parec |
Supported |
Note: WSL2 is not supported. RDPSink cannot capture Windows system audio.
PipeWire (Ubuntu 24.04, real hardware):
PipeWire 1.0.5
Server Name: PulseAudio (on PipeWire 1.0.5)
pw-record --target 50 → captured and verified with audio playback
- Linux (Ubuntu 24.04 recommended)
- Python 3.6+
- PipeWire (
pw-record) or PulseAudio (parec,pactl)
MIT