ghostmeet for Openclaw

A privacy-first AI meeting assistant that provides local Whisper transcription and Claude-powered summaries for browser-based meetings.

higangssh
v1.0.4
Mar 8, 2026
1
1.1k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install ghostmeet

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install ghostmeet using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is ghostmeet?

ghostmeet is a robust, self-hosted AI meeting assistant that enables users to capture, transcribe, and summarize discussions directly from their browser. By utilizing local Whisper models, it ensures that sensitive audio data remains on your machine, providing a secure alternative to cloud-heavy meeting bots. This integration via Openclaw Skills allows you to interact with your meeting history through a seamless chat interface.

The skill connects to a Docker-based backend and a Chrome extension to bridge the gap between live audio and actionable text. With the power of Openclaw Skills, developers can programmatically retrieve transcripts, check system health, and generate high-level summaries that highlight key decisions and action items using the Anthropic Claude API.

ghostmeet Use Cases

  • Automatically generating post-meeting action items and summaries for team syncs.
  • Searching through historical meeting transcripts to find specific project decisions.
  • Monitoring meeting frequency and duration through session listing and metadata analysis.
  • Maintaining a private, local archive of all browser-based communications without third-party data access.

How ghostmeet Works

  1. The user starts a meeting in a Chrome tab and activates the ghostmeet extension to begin capturing audio.
  2. A local Docker container running Whisper processes the audio in real-time to create timestamped transcript segments.
  3. The Openclaw Skills interface queries the backend to list recent sessions and identify the correct meeting ID.
  4. Upon request, the skill retrieves the full text transcript or sends it to the Claude API for intelligent summarization.
  5. The final summary, including next steps and key takeaways, is formatted and presented to the user.

ghostmeet Setup

To get started, you must first set up the ghostmeet backend using Docker:

git clone https://github.com/Higangssh/ghostmeet.git
cd ghostmeet
cp .env.example .env
# Set GHOSTMEET_ANTHROPIC_KEY in .env for summary features
docker compose up -d

After the backend is running, install the Chrome Extension from the extension/ folder in developer mode. Ensure your Openclaw Skills environment can reach the default backend at http://127.0.0.1:8877.

ghostmeet Data Schema & Taxonomy

The skill manages data through a structured API that organizes information into sessions and segments:

Object Description
Session ID Date-based unique identifier (e.g., 20260308-065021).
Transcript A collection of segments containing start/end timestamps and transcribed text.
Summary AI-generated text block containing decisions, action items, and next steps.
Health Data System metadata including the active Whisper model (tiny, base, small, etc.) and device type.

ghostmeet Advanced Features

  • Local-first transcription ensures maximum privacy as audio never leaves your local network.
  • Support for multiple Whisper model sizes (from tiny to large-v3) to balance speed and accuracy.
  • Configurable chunk intervals and language auto-detection for optimized transcription performance.
  • Seamless integration with Openclaw Skills for automated post-meeting documentation workflows.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*