Quick start

Install the Gemini Live Discord voice bridge, wire SORA bridge elements if needed, restart Hermes, then start /voice-live from Discord.

Install

git clone https://github.com/Capslockb/hermes-live-discord-agent-plugin.git
cd hermes-live-discord-agent-plugin
cd installer
./install.py
# or, from the repo checkout:
./install.sh --from-local

cd ..
python3 installer/enable_sora_bridge_elements.py
python3 -m py_compile plugin/sora_bridge_elements.py plugin/__init__.py
systemctl --user restart hermes-gateway

First session

From Discord, join a voice channel, then run:

/voice-live          # join
/voice-live-leave    # leave

Verify

curl -s http://127.0.0.1:18943/health | python3 -m json.tool

You should see bridge health after /voice-live has started a session. If the bridge is not running yet, the sidecar health request can fail normally.

Verify SORA bridge elements

sora_bridge_preflight
sora_redact text="Authorization: Bearer fake.fake.fake"
sora_live_grill text="migrate SORA bridge features into Gemini bridge"
sora_goal_synth text="migrate SORA bridge features into Gemini bridge"

Common pitfalls

• /voice-live cannot infer your channel — join a voice channel first with the configured Discord user.
• /health fails — start a session first or check DISCORD_VOICE_LIVE_PORT.
• SORA tools missing — run installer/enable_sora_bridge_elements.py and restart Hermes.
• The model cannot see screenshare — use a screenshot, voice_live_frame, or the frame feeder.
• Optional tools fail — mail, GitHub, Spotify, Home Assistant, and CLI delegation require their own credentials/install.