Last month, I asked ChatGPT to help me update a project tracker in Notion based on emails it had just analyzed. It understood exactly what I needed. It drafted the perfect database entries. Then it told me to copy and paste everything manually.
That's when I realized the problem with most AI assistants: they can see your work, but they can't touch it. They're brilliant consultants who aren't allowed to write anything down.
Claude changed that for me. Through something called the Model Context Protocol (MCP), Claude doesn't just read my Notion workspace—it creates pages, updates databases, and manages content directly. The same AI conversation that analyzes my project requirements can also build the documentation. No copying. No pasting. No context switching.
This isn't a minor convenience. It's the difference between an AI that advises and an AI that executes. As of December 2025, MCP has become the industry standard, with OpenAI, Google, and Microsoft adopting it through the Linux Foundation's new Agentic AI Foundation.
Why Your Microsoft Copilot Licenses Are Collecting Dust & How to Fix
This workshop is for leaders, managers, and professionals who: Rolled out Copilot and saw adoption stall. Have teams with "access" but aren't getting value, and are paying for unused seats. Tried group training and watched it fade after two weeks. Need to show ROI on AI investments.
Bozhanka "Boz" Vitanova brings deep expertise in AI implementation, systems thinking, and skills development. She started TeamLift to transform how people grow through disruption, not by resisting change, but by building through it. Previously, Boz served as a National Science Foundation I-Corps Instructor at Brandeis University, where she coached PhD researchers on turning academic ideas into real-world impact.
She co-founded EML Solutions, using skills data to power project-based teaming at companies like Unilever and Philips, and Yunus&Youth, a global social enterprise supported by Nobel Peace Laureate Muhammad Yunus. A Fulbright Scholar, WEF Global Shaper, and One Young World Ambassador, Boz blends technical fluency with a deep understanding of human potential to design learning systems that help people rise, even as everything shifts.
Join us on January, 20, 12:00 PM EST and leave with practical frameworks you can use immediately!
Why I Switched from ChatGPT to Claude for Real Work
The Model Context Protocol gives AI the hands to match its eyes
MCP is an open standard that creates secure, two-way connections between AI assistants and external tools. Think of it as USB-C for AI—a universal interface that lets any compliant AI system read from and write to any compliant application.
The protocol operates on three core primitives: Resources (read-only data access), Tools (action execution), and Prompts (contextual templates). This architecture enables Claude to query your Notion database, analyze the results, and create new pages based on that analysis—all in a single conversation.
I am using Notion in my example, but this could be Google Workspace, Microsoft, or myriad other applications. What makes this interesting is that you can use Claude or ChatGPT more like a desktop app, calling those tools and apps directly through the interface.
The more you integrate, the more you can use your ChatGPT interface or Claude Interface in this example to orchestrate other apps without leaving your AI Chatbot Interface.
Using Claude to Write Updates to My Book Manuscript in Notion
How to Access MCP Connectors in Claude
Requirements:
Claude Pro, Team, or Enterprise subscription ($20+/month)
Claude.ai web interface or Claude Desktop app
Web Access:
Step 1: Open Claude.ai and start a new conversation
Step 2: Click the tools icon (wrench) in the chat input area
Step 3: Select "Add connectors" to open the directory
Step 4: Browse available connectors under the "Web" or "Desktop extensions" tabs
Step 5: Click the plus (+) icon next to your chosen tool and complete OAuth authentication
The connector directory is also accessible directly at claude.ai/settings/connectors or through Anthropic's directory announcement page.
Current connector count: 75+ pre-built integrations as of January 2025, with more added regularly. Browse the complete list at the official MCP Registry.
Claude’s Settings for Integrating Connectors which are connected via MCP
Notion Integration: Read AND Write Access
I primarily write to two systems: Google Workspace and Notion. Here’s how it works in Notion; the process is similar for most systems you may integrate with.
What it does: Connects Claude to your Notion workspace with full read-write permissions, allowing natural language commands to search, create, update, and manage pages and databases.
Step 1: Navigate to Claude's connector directory and select "Notion MCP"
Step 2: Authorize access through Notion's OAuth flow
Step 3: Select which workspace(s) to connect
Step 4: Start using natural language commands
Example commands:
"Search my Notion for all project pages mentioning 'Q1 launch'"
"Create a new page in my Projects database with title 'Website Redesign' and status 'Planning'"
"Update the 'Marketing Campaign' page to change the deadline to February 15"
"Build a book tracker database with book and author tables linked by a two-way relation"
Time: Initial setup takes 2-3 minutes. Commands execute in seconds.
Limitations:
Claude operates with your Notion permissions—it can access everything you can access
Semantic search isn't perfect; use specific terms rather than vague descriptions
Complex database relations may require multiple commands to build correctly
For detailed setup instructions, see Notion's MCP documentation.
Google Drive and Gmail: Context Without Copy-Paste
What it does: Lets Claude search, read, and reference documents from your Google Drive and emails from Gmail within conversations.
Practical workflows:
Research synthesis: "Find all documents in my Drive mentioning 'competitive analysis' and summarize the key findings"
Email context: "Check my Gmail for the thread with Sarah about the budget proposal and help me draft a follow-up"
Cross-reference: "Compare what's in my Q3 report document with the data from last week's finance team emails"
Limitations:
Write access to Google services requires additional automation tools like Zapier's MCP integration
Large documents may need to be processed in sections
Shared drives require appropriate permissions
Multi-Tool Workflows: Where MCP Shines
The real power of MCP emerges when you chain multiple tools together. Here's a workflow I use weekly:
"Check my calendar for meetings this week, pull the associated
project pages from Notion, and draft a status update email
summarizing progress on each initiative"
Claude accesses three systems (Calendar, Notion, Gmail), synthesizes information across them, and produces an actionable output—all from a single prompt.
Other high-value combinations:
Linear + Notion: Generate release notes from sprint tickets
Asana + Google Docs: Create project briefs from task lists
Canva + Notion: Design assets informed by brand guidelines stored in your workspace
For more workflow ideas, see Anthropic's guide to extending Claude's capabilities.
Claude’s Director of Connectors
What MCP Can't Do (Yet)
Real-time collaboration: MCP connections are request-response, not continuous. Claude won't automatically update as your tools change.
Complex approval workflows: Actions execute immediately once you approve. There's no built-in staging or review process for multi-step changes.
Universal write access: Some integrations remain read-only. Google Drive and Gmail, for example, still require Zapier or similar tools for write operations.
Offline access: MCP requires active connections. If a service is down, Claude can't access it.
Deep semantic understanding: Claude can search your content, but it's not truly understanding relationships between documents the way a human familiar with your work would.
The Industry Shift: Why This Matters Now
In December 2025, Anthropic donated MCP to the newly formed Agentic AI Foundation under the Linux Foundation. Co-founders include OpenAI, Block, Google, Microsoft, AWS, and Cloudflare.
This isn't just a technical milestone—it's a signal that the fragmented era of AI integrations is ending. The MCP SDK has reached 97 million monthly downloads with over 10,000 public servers in operation.
For business users, this means:
Connectors you set up today will work with future AI tools
Your automation investments won't be stranded by platform changes
The ecosystem of available integrations will only grow
For protocol updates and new features, follow the MCP Blog.
Getting Started Today
Right now: Connect Notion to Claude using the directory at claude.ai/settings/connectors. Try a simple search command to verify the connection works.
This week: Identify your most copy-paste-heavy workflow. Build a multi-step prompt that lets Claude handle the data transfer.
Next week: Add a second connector and experiment with cross-tool commands. Calendar + Notion or Gmail + Drive are good starting combinations.
This month: Document what's working. MCP workflows compound—each successful automation reveals two more opportunities.
The shift from AI-as-advisor to AI-as-collaborator is happening now. The tools are live, the standard is set, and the only variable is whether you'll integrate them before the efficiency gap with AI-native competitors becomes a problem.
This is where AI gets real. No sales pitches—just 4,000 builders, engineers, operators, and execs sharing how AI actually works in practice, from hands-on workshops to real-world sessions, right in the heart of RTP and Duke’s AI ecosystem.
We want to make learning AI as practical, accessible, and immediately useful as possible—so you leave knowing what to build, what to deploy, and what actually drives results.
I appreciate your support.
Your AI Sherpa,
Mark R. Hinkle
Publisher, The AIE Network
Connect with me on LinkedIn
Follow Me on Twitter
