Package Exports

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (recap.js) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

Recap.js

Record. Transcribe. Summarize.

🎤 A powerful CLI tool and TypeScript library for automated meeting notes, powered by OpenAI Whisper and GPT models.

✨ Features

🎙️ Record audio from your microphone with customizable settings
🎯 Transcribe speech to text using OpenAI Whisper or Google Gemini AI
📝 AI-powered summarization with multiple providers (OpenAI GPT, Google Gemini)
🔄 Multiple AI providers - Choose between OpenAI and Gemini or let Recap.js decide
📊 Provider comparison - Compare results from different AI providers side-by-side
📄 Multiple export formats (Markdown, JSON, Plain Text)
🔄 Complete workflows with a single command
💻 CLI and programmatic APIs for maximum flexibility
🎨 Beautiful terminal UI with progress indicators and colors
⚡ TypeScript-first with full type safety

🚀 Quick Start

Installation

npm install -g recap.js

Setup (Interactive Tutorial)

recapjs setup

This will:

Show you exactly how to get API keys
Guide you through the setup process step-by-step
Save your preferences for daily use
Test your configuration

Alternative: Manual Setup

# OpenAI (Higher quality, ~$0.006/minute)
export OPENAI_API_KEY="your-openai-api-key-here"

# Google Gemini (More cost-effective, ~$0.002/minute)
export GEMINI_API_KEY="your-gemini-api-key-here"

💡 Tip: Start with recapjs setup - it includes a helpful tutorial for getting API keys!

Basic Usage

# First-time setup (saves your preferences)
recapjs setup

# Quick record with saved settings
recapjs quick-record --duration 300

# Traditional commands
recapjs process --duration 300 --provider gemini --output meeting-notes
recapjs transcribe meeting.wav --provider gemini --format json
recapjs summarize transcript.json --provider openai --style action-items

# Compare providers on the same audio
recapjs compare meeting.wav --style meeting-notes

# Configuration management
recapjs config show
recapjs config set provider gemini
recapjs config set style action-items

# Check provider status
recapjs providers

# Get audio file information
recapjs info meeting.wav

📖 CLI Commands

`recapjs record`

Record audio from your microphone.

recapjs record [options]

Options:
  -o, --output <path>     Output file path (default: "recording.wav")
  -d, --duration <sec>    Recording duration in seconds
  -s, --sample-rate <hz>  Sample rate (default: 16000)
  --summarize            Auto-transcribe and summarize after recording
  --format <format>      Export format: markdown|json|txt (default: "markdown")
  --style <style>        Summary style (default: "meeting-notes")
  --provider <provider>  AI provider: openai|gemini|auto (default: "auto")
  --model <model>        Specific model to use

Examples:

# Record for 5 minutes
recapjs record --duration 300 --output team-standup.wav

# Record and auto-summarize with action items using Gemini
recapjs record --duration 600 --summarize --style action-items --provider gemini

`recapjs transcribe`

Transcribe audio files to text.

recapjs transcribe <audio-file> [options]

Options:
  -o, --output <path>     Output file path
  --format <format>       Output format: json|txt|markdown (default: "json")
  --language <lang>       Audio language (e.g., en, es, fr)
  --provider <provider>   AI provider: openai|gemini|auto (default: "auto")
  --model <model>         Model to use (whisper-1, gemini-1.5-pro, etc.)

Examples:

# Transcribe to JSON with OpenAI Whisper
recapjs transcribe meeting.mp3 --provider openai --format json

# Transcribe Spanish audio with Gemini
recapjs transcribe reunion.wav --provider gemini --language es --format markdown

`recapjs summarize`

Generate AI-powered summaries from transcripts.

recapjs summarize <input-file> [options]

Options:
  -o, --output <path>     Output file path
  --style <style>         Summary style (default: "meeting-notes")
  --format <format>       Output format: markdown|json|txt (default: "markdown")
  --provider <provider>   AI provider: openai|gemini|auto (default: "auto")
  --model <model>         Model to use (gpt-3.5-turbo, gpt-4, gemini-1.5-pro, etc.)

Summary Styles:

action-items - Focus on tasks and assignments
meeting-notes - Traditional meeting minutes format
bullet-points - Concise bullet-point summary
executive-summary - High-level overview for leadership
detailed - Comprehensive analysis with context

Examples:

# Generate action items with Gemini
recapjs summarize transcript.json --provider gemini --style action-items

# Executive summary with GPT-4
recapjs summarize meeting.txt --provider openai --style executive-summary --model gpt-4

# Compare providers on the same transcript
recapjs compare meeting.wav --style executive-summary

`recapjs process`

Complete workflow: record → transcribe → summarize.

recapjs process [options]

Options:
  -o, --output <base>     Output file base name (default: "meeting")
  -d, --duration <sec>    Recording duration in seconds
  --style <style>         Summary style (default: "meeting-notes")
  --format <format>       Final output format (default: "markdown")
  --provider <provider>   AI provider: openai|gemini|auto (default: "auto")
  --model <model>         Model to use
  --keep-intermediate     Keep audio and transcript files

Examples:

# Complete 10-minute meeting workflow with Gemini
recapjs process --duration 600 --provider gemini --output board-meeting

# Generate action items and keep all files
recapjs process --style action-items --keep-intermediate

# Use specific models
recapjs process --provider openai --model gpt-4 --duration 300

`recapjs compare`

Compare AI providers on the same audio file.

recapjs compare <audio-file> [options]

Options:
  --style <style>         Summary style for comparison (default: "meeting-notes")
  --output <path>         Output directory for results (default: "./comparison-results")

Examples:

# Compare providers with meeting notes style
recapjs compare meeting.wav --style meeting-notes

# Compare with custom output directory
recapjs compare meeting.mp3 --output ./provider-comparison

`recapjs providers`

Show available AI providers and their configuration status.

recapjs providers

This command will show:

✅ Connected providers with available models
❌ Providers with connection issues
⚠️ Providers missing API keys
💡 Configuration tips

`recapjs setup`

Run the initial configuration wizard.

recapjs setup

Walks you through setting up API keys, default provider, export format, and summary style.

`recapjs config`

Manage persistent configuration settings.

recapjs config show                    # Display current configuration
recapjs config set <key> <value>       # Set configuration value
recapjs config clear                   # Clear all configuration

Available config keys:

openai-key - OpenAI API key
gemini-key - Gemini API key
provider - Default AI provider (openai|gemini)
format - Default export format (markdown|json|txt)
style - Default summary style
model - Default AI model

`recapjs quick-record`

Quick record using your saved configuration preferences.

recapjs quick-record [options]

Options:
  -d, --duration <seconds>    Recording duration in seconds
  -o, --output <path>         Output file path

This command uses your saved defaults for provider, format, and style.

`recapjs info`

Show detailed information about audio files.

recapjs info <audio-file>

🛠️ Programmatic API

Use Recap.js in your Node.js applications:

import {
  AudioRecorder,
  UnifiedAIProvider,
  GeminiTranscriber,
  GeminiSummarizer,
  DataExporter,
  createRecapWorkflow,
} from 'recap.js';

// Complete workflow with multiple providers
const result = await createRecapWorkflow({
  openaiApiKey: 'your-openai-key',
  geminiApiKey: 'your-gemini-key',
  provider: 'gemini', // or 'openai' or 'auto'
  recordingOptions: {
    outputPath: 'meeting.wav',
    duration: 300, // 5 minutes
    sampleRate: 16000,
  },
  summaryStyle: 'action-items',
  exportFormat: 'markdown',
  exportPath: 'meeting-summary.md',
});

if (result.success) {
  console.log('Summary saved to:', result.data.exportPath);
  console.log('Action items:', result.data.summary.action_items);
}

// Using the unified AI provider
const aiProvider = new UnifiedAIProvider('your-openai-key', 'your-gemini-key');

const recorder = new AudioRecorder();
const exporter = new DataExporter();

// Record audio
const recording = await recorder.startRecording({
  outputPath: 'audio.wav',
  duration: 180,
});

// Transcribe with auto provider selection
const transcript = await aiProvider.transcribeAudio({
  inputPath: 'audio.wav',
  provider: 'auto', // Let Recap.js choose the best provider
});

// Summarize with specific provider
const summary = await aiProvider.summarizeFromTranscript(transcript.data!, {
  style: 'meeting-notes',
  provider: 'gemini',
});

// Compare providers
const comparison = await aiProvider.compareProviders({
  inputPath: 'audio.wav',
  style: 'action-items',
});

if (comparison.success) {
  console.log('OpenAI results:', comparison.data.openai);
  console.log('Gemini results:', comparison.data.gemini);
}

// Export
await exporter.exportData({
  data: summary.data!,
  format: 'markdown',
  outputPath: 'summary.md',
});

📋 Example Output

Meeting Summary (Markdown)

# Weekly Team Standup

**Date:** 2024-01-15
**Duration:** 23 minutes
**Summary Style:** meeting-notes

## Participants

- **Alice Johnson** (Team Lead)
  - Discussed project timeline updates
  - Reviewed Q1 objectives
- **Bob Smith** (Developer)
  - Reported on API development progress
- **Carol Williams** (Designer)
  - Presented new UI mockups

## Key Topics

- Q1 project roadmap review
- API development status
- UI/UX design updates
- Resource allocation for February

## Summary

The team reviewed progress on Q1 objectives and discussed upcoming deliverables. API development is on track with the authentication module completed. New UI designs were well-received and approved for implementation.

## Action Items

### High Priority

- [ ] **Complete user authentication API endpoints**
  - **Assignee:** Bob Smith
  - **Deadline:** January 20th

- [ ] **Finalize UI component library**
  - **Assignee:** Carol Williams
  - **Deadline:** January 25th

### Medium Priority

- [ ] **Update project documentation**
  - **Assignee:** Alice Johnson
  - **Deadline:** End of month

## Decisions Made

1. Approved new UI design direction for user dashboard
2. Decided to prioritize mobile responsiveness in February
3. Agreed to weekly design reviews starting next week

## Next Steps

1. Schedule design review sessions for next week
2. Begin mobile optimization planning
3. Update stakeholders on Q1 progress

---

_Generated by Recap.js on 1/15/2024, 2:30:15 PM_

⚙️ Configuration

Environment Variables

# AI Provider API Keys (at least one required)
OPENAI_API_KEY=your-openai-api-key
GEMINI_API_KEY=your-gemini-api-key
# or
GOOGLE_API_KEY=your-google-api-key

# Optional Configuration
RECAP_DEFAULT_PROVIDER=auto  # auto|openai|gemini
RECAP_DEFAULT_MODEL=auto     # or specific model names
RECAP_DEFAULT_FORMAT=markdown
RECAP_DEFAULT_STYLE=meeting-notes

Global CLI Options

# Set API keys via command line
recapjs --openai-key YOUR_KEY --gemini-key YOUR_KEY [command]

# Choose provider globally
recapjs --provider gemini [command]

# Check current configuration
recapjs providers

Audio Requirements

Supported formats: WAV, MP3, M4A, FLAC, OGG, WebM, MP4
Maximum file size: 25MB (OpenAI API limit)
Recommended settings: 16kHz sample rate, mono channel
Recording requirements: Microphone access, SoX for recording

💰 Cost Estimation

Recap.js supports multiple AI providers with different pricing:

OpenAI Pricing

Whisper (transcription): $0.006 per minute
GPT-3.5-turbo (summarization): ~$0.001-0.003 per meeting
GPT-4 (summarization): ~$0.01-0.03 per meeting

Google Gemini Pricing

Gemini 1.5 Pro (transcription): ~$0.002 per minute
Gemini 1.5 Pro (summarization): ~$0.0005-0.002 per meeting
Gemini 1.5 Flash (faster/cheaper): ~$0.0001-0.0005 per meeting

Example costs for a 30-minute meeting:

OpenAI (Whisper + GPT-3.5-turbo):

Transcription: $0.18
Summarization: ~$0.002
Total: ~$0.18

Gemini (1.5 Pro):

Transcription: ~$0.06
Summarization: ~$0.001
Total: ~$0.06

Compare costs:

recapjs info <audio-file>  # Shows estimated costs for all providers
recapjs compare <audio-file>  # Compare quality and costs side-by-side

🔧 Development

Prerequisites

Node.js 16+
TypeScript 5.2+
SoX (for audio recording)

Setup

# Clone repository
git clone https://github.com/rollenasistores/recap.js.git
cd recap.js

# Install dependencies
npm install

# Build project
npm run build

# Run in development
npm run dev

# Run linting
npm run lint

# Format code
npm run format

Project Structure

src/
├── cli.ts          # CLI entry point
├── index.ts        # Library exports
├── types.ts        # TypeScript interfaces
├── recorder.ts     # Audio recording
├── transcriber.ts  # Speech-to-text
├── summarizer.ts   # AI summarization
├── exporter.ts     # Export functionality
└── utils.ts        # Utility functions

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

Fork the repository
Create a feature branch: git checkout -b feature/amazing-feature
Commit changes: git commit -m 'Add amazing feature'
Push to branch: git push origin feature/amazing-feature
Open a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🆘 Support

🚀 Roadmap

Speaker diarization (identify who said what)
Real-time transcription and summarization
Integration with calendar apps
Custom AI prompts and templates
Multi-language support enhancements
Web interface for non-CLI users
Integration with popular meeting platforms

Made with ❤️ for better meetings
_{Record. Transcribe. Summarize.}