Claunnector

Product Introduction

1. Definition: Claunnector is a native macOS menubar application and local Model Context Protocol (MCP) server. It serves as a secure bridge that provides Large Language Models (LLMs) and AI agents—including Claude Desktop, ChatGPT, and Codex—direct programmatic access to local macOS system services such as Mail, Calendars, Reminders, Contacts, and Notes.

2. Core Value Proposition: Claunnector exists to solve the "context gap" between powerful AI models and a user's private data without compromising security. By leveraging the Model Context Protocol (MCP), it enables users to turn their local Mac into an AI-augmented workstation. The primary keyword-driven value is providing local AI data access and secure MCP tool integration, ensuring that sensitive personal information never leaves the machine while allowing AI tools to perform complex tasks like drafting emails, scheduling events, and managing contacts natively.

Main Features

1. 70+ Integrated MCP Tools: Claunnector offers a comprehensive suite of over 70 tools spanning five core macOS services. This includes a full-featured Mail module supporting Search, Read, Send, Reply, and Move operations with Markdown-to-HTML conversion and SPF/DKIM/DMARC authentication. The Calendar and Reminders modules utilize macOS EventKit for full CRUD (Create, Read, Update, Delete) capabilities, including support for recurrence rules and alarms. The Contacts module uses the Contacts.framework for organizing iCloud and Google contacts, while the Notes module allows for deep manipulation of Apple Notes folders and content.

2. Security-First Architecture (Localhost & Sandbox): The application is engineered with a strict privacy-by-design philosophy. The MCP server binds exclusively to the 127.0.0.1 (localhost) loopback interface, making remote connections physically impossible. It operates within a macOS App Sandbox with minimal necessary entitlements and stores all sensitive credentials, such as Gmail OAuth tokens and IMAP passwords, within the secure macOS Keychain via the Security framework.

3. Granular Client Profiles and Token Authorization: Users can create specific "Client Profiles" to enforce per-client access control. This allows a user to grant a "Calendar Bot" access only to specific calendars while preventing it from reading emails. Each profile is issued a unique bearer token for authorization, ensuring that AI clients like Claude Code or ChatGPT only interact with the specific modules and resources permitted by the user.

4. 1-Click Installation and Multi-Protocol Support: Claunnector simplifies the complex setup of AI tools by providing one-click configuration for Claude Desktop, Claude Code, and Claude Cowork. It supports multiple communication protocols, including stdio for local process communication and HTTP+SSE (Server-Sent Events) for web-based or remote-compatible clients like ChatGPT and Codex, allowing any MCP-compliant client to connect via a standard URL (http://127.0.0.1:5000/sse).

5. Full Audit Log and Native Notifications: To ensure transparency, the app maintains a comprehensive Audit Log that records every tool call with timestamps, client profile details, and specific module usage. Additionally, AI tools can trigger native macOS notification banners to inform the user of completed actions, such as "Meeting Scheduled" or "Email Sent," maintaining a history of AI interactions within the app.

Problems Solved

1. Privacy Risks of Cloud-Based AI Integration: Many AI integrations require syncing local data to third-party clouds. Claunnector solves this by keeping all data processing on the user's Mac. This addresses the needs of users who require local AI privacy and secure data handling for sensitive professional workflows.

2. Lack of Context for AI Assistants: Standard AI models lack "knowledge" of a user's current schedule, unread emails, or contact lists. Claunnector provides the necessary Model Context Protocol infrastructure to ground AI responses in real-world personal data, making the AI significantly more useful for administrative and productivity tasks.

3. Configuration Complexity for Developers: Setting up MCP servers manually often involves editing JSON config files and managing background daemons. Claunnector solves this with a GUI-based menubar app that handles all configuration, daemon management, and protocol bridging automatically.

4. Target Audience: The product is designed for "Power Users" who utilize AI for daily productivity, "Software Developers" using Claude Code for project management and communication, "Privacy-Conscious Professionals" (Lawyers, Executives) who cannot upload data to the cloud, and "AI Enthusiasts" looking to build custom local agentic workflows on macOS.

5. Use Cases: Scenarios include asking Claude to "Summarize my unread emails from today," requesting ChatGPT to "Find a free 30-minute slot on my calendar tomorrow and invite John," or using Claude Code to "Send a status update email to the team based on my latest notes."

Unique Advantages

1. Native macOS Optimization: Unlike generic Node.js or Python MCP servers, Claunnector is a native macOS application built specifically for the platform. This ensures better performance, lower resource consumption, and seamless integration with macOS-specific frameworks like EventKit and the Apple Notes database.

2. Zero-Configuration Bridge: The built-in stdio bridge for Claude Desktop is a significant differentiator. It removes the need for users to manually locate and edit claude_desktop_config.json, which is a common friction point in the MCP ecosystem.

3. Transparency and Accountability: The searchable Audit Log is a unique security feature that provides a "black box" record of AI behavior. Users can verify exactly what their AI has done, providing a layer of safety not found in most opaque AI integration tools.

Frequently Asked Questions (FAQ)

1. Is Claunnector secure and does my data leave my Mac? No data ever leaves your Mac. Claunnector runs entirely on localhost (127.0.0.1), which means the server is not accessible from the internet. All processing happens locally, and your passwords/tokens are stored in the encrypted macOS Keychain.

2. How do I connect Claude Desktop to Claunnector? Claunnector features a 1-click install for Claude Desktop. In the application settings, simply click the "Install" button, and Claunnector will automatically write the necessary configuration to your Claude Desktop settings file. Once restarted, Claude will have access to the tools you have enabled in your profile.

3. Can I use Claunnector with ChatGPT or other LLMs? Yes. Claunnector supports the HTTP+SSE protocol. Any MCP-compatible client can connect by pointing to the local URL (usually http://127.0.0.1:5000/sse). This makes it compatible with ChatGPT, Codex, and other emerging AI tools that support the Model Context Protocol.

4. What versions of macOS are supported? Claunnector requires macOS 15 or later to take advantage of the latest system frameworks and security features required for seamless integration with Mail, Contacts, and Calendars.