Files
hermes-agent/plugin-catalog/gemini-live-bridge.yaml

28 lines
1.4 KiB
YAML

name: gemini-live-bridge
repo: https://github.com/cruzlxyz/gemini-live-bridge
sha: a0016e33244ff2f74fd33bae1869e4a8c06f5034
description: >-
Realtime duplex voice chat with the Gemini Live API — a status-bar minibar for the Hermes
desktop app (transcript, model catalog, thinking level, session token meter, auto-redial via
session resumption). Full-proxy architecture: the plugin backend owns the upstream Live
WebSocket and the user's key (requires_env GOOGLE_API_KEY); the desktop UI only talks to
localhost and receives single-use ephemeral tokens. No agent tools; desktop-focused, dashboard
tab hidden. Disclosure — pressing Start captures microphone audio and streams it through the
local backend to generativelanguage.googleapis.com using GOOGLE_API_KEY (or GEMINI_API_KEY) from
the environment; input/output transcripts are held in backend memory for the session, and a
Gemini session-resumption handle is written to .resume.json inside the plugin's install
directory.
maintainer: cruzlxyz
tier: community
version: "1.0.6"
category: voice
docs_url: https://github.com/cruzlxyz/gemini-live-bridge#readme
image: https://raw.githubusercontent.com/cruzlxyz/gemini-live-bridge/a0016e33244ff2f74fd33bae1869e4a8c06f5034/docs/banner.png
platforms: []
capabilities:
provides_tools: []
provides_hooks: []
provides_middleware: []
requires_env:
- GOOGLE_API_KEY