We're thrilled to announce Clueless v1.0.0, our first stable release! This major milestone brings a complete architectural overhaul, enhanced performance, and production-ready features for AI-powered meeting assistance.

🌟 Highlights

• Production-Ready Architecture: Complete rewrite with modular component architecture and state management
• Official OpenAI SDK Integration: Migrated to OpenAI Agents SDK for improved reliability
• Enhanced macOS Support: Native permissions handling for microphone and screen recording
• Real-time Data Persistence: Automatic conversation saving with live transcription storage
• Improved User Experience: Streamlined onboarding and cleaner interface

✨ New Features

OpenAI Agents SDK Integration

• Replaced custom WebSocket implementation with official @openai/agents and @openai/agents-realtime SDKs
• Improved connection stability and error handling
• Better TypeScript support with official type definitions
• Maintained dual-agent architecture (salesperson + customer coach)

macOS Permissions System

• Integrated node-mac-permissions for proper permission handling
• Added microphone, screen recording, and camera permission management
• Context-isolated Electron implementation for enhanced security
• Visual permission status indicators in the UI

System Audio Capture

• Integrated electron-audio-loopback for reliable system audio capture
• Eliminated dependency on Swift-based audio capture
• Cross-platform compatibility improvements
• Real-time audio level monitoring for both microphone and system audio

Data Persistence

• Automatic conversation session creation on call start
• Real-time transcript saving with 5-second intervals
• Comprehensive insight, topic, commitment, and action item tracking
• Call history with detailed conversation analysis

Developer Experience

• Mock data system for UI testing without API calls
• Comprehensive error handling and logging
• Build system improvements for NativePHP distribution
• GitHub Actions CI/CD pipeline fixes

🔧 Improvements

Architecture & Performance

• Component-Based Architecture: Migrated from monolithic Main.vue (1,558 lines) to 14+ modular components
• State Management: Implemented 3 Pinia stores for business logic, settings, and OpenAI SDK management
• Memory Optimization: Fixed memory leaks in audio capture and WebSocket connections
• Build Performance: Optimized frontend asset building for production

User Interface

• Streamlined Navigation: Removed redundant navigation items and badges
• Modal Onboarding: Replaced dedicated onboarding page with inline modal
• Connection Status: Improved visual feedback with color-coded states
• Responsive Design: Enhanced mobile layout with proper card sizing
• Dark Mode: Consistent theming across all components

Security & Permissions

• Context Isolation: Implemented secure IPC communication for Electron
• Permission Handling: Graceful degradation when permissions are denied
• API Key Management: Secure storage with cache-based implementation

🐛 Bug Fixes

• Fixed duplicate template seeding on app startup
• Resolved npm optional dependencies issue in CI/CD pipeline
• Fixed conversation saving in RealtimeAgent v2
• Eliminated double scrollbar in main interface
• Fixed dropdown z-index issues with Teleport solution
• Resolved ESLint errors across the codebase
• Fixed Vite hot reload issues in production builds
• Corrected WebSocket parameter names for OpenAI API

💔 Breaking Changes

• Removed standalone Onboarding.vue page component (replaced with modal)
• Removed CheckOnboarding middleware
• Updated navigation routes to point to /realtime-agent-v2
• Minimum macOS version requirement for system audio features

📦 Dependencies

• Added: @openai/agents (^0.0.12), @openai/agents-realtime (^0.0.12)
• Added: electron-audio-loopback (^1.0.5)
• Added: node-mac-permissions (^2.5.0) - optional dependency
• Updated: Vue to 3.5.13, Vite to 6.2.0, TypeScript to 5.2.2

🔄 Migration Guide

For users upgrading from pre-release versions:

1. Clear application cache to ensure clean state
2. Re-grant permissions for microphone and screen recording
3. Update API keys through Settings if needed
4. Existing conversations are preserved and compatible

🙏 Acknowledgments

Special thanks to all contributors and testers who helped make this stable release possible. This release includes contributions from both human developers and AI assistance through Claude.

📋 Technical Details

• 63 files changed with 9,445 additions and 2,296 deletions
• 30+ commits addressing various features and fixes
• Comprehensive test coverage for critical functionality
• Production-ready build pipeline with NativePHP

Download: Available for macOS (Intel and Apple Silicon)
Requirements: macOS 10.15+, 4GB RAM minimum
License: MIT with Commons Clause

AI-Powered Meeting Assistant for macOS

We're excited to announce the first beta release of Clueless, your intelligent conversation companion that provides real-time transcription, analysis, and insights for all your meetings and calls.

🎯 Core Features

Real-time Conversation Intelligence

• Live voice transcription powered by OpenAI's Realtime API
• Instant AI-generated insights and summaries
• Automatic action item extraction
• Speaker identification and sentiment analysis

Smart Coaching Templates

• Pre-built AI coaching personas for different meeting types
• Create custom templates with dynamic variables
• Real-time guidance during conversations
• Template categories for easy organization

Powerful Analysis

• Topic detection and categorization
• Commitment and promise tracking
• Sales stage analysis and intent classification
• Cross-conversation context retention

Desktop-First Experience

• Native macOS application
• Overlay mode for floating above other apps
• System tray integration
• Keyboard shortcuts for quick actions
• Dark/light theme support

Privacy & Data Control

• All data stored locally on your device
• No authentication or cloud storage required
• Screen protection mode for sensitive conversations
• Full control over API key management

🚀 Getting Started

1. Download and install Clueless
2. Add your OpenAI API key in Settings
3. Start your first conversation!

⚠️ Beta Notes

This is a beta release. While core features are stable, you may encounter occasional issues. Please report any bugs or feedback to help us improve.

📋 Requirements

• macOS 10.15 or later
• OpenAI API key with Realtime API access
• Active internet connection for transcription