The problem: AI tools work side-by-side, not together

The reality in many companies today looks like this: the marketing team develops campaign ideas in Claude, developers code with Cursor, the project team coordinates in Slack, and the actual plans, timelines, and approvals reside in a separate work tool like monday.com. Each of these tools runs independently, with its own context.

The result: Insights from AI chats remain stuck in those very chats. Results must be manually transferred from one tool to the next, and context is lost with every switch.

This is also one of the reasons why many companies have yet to feel the true value of their AI investments: the technology makes individuals more productive, but does little for team collaboration as long as it remains siloed. This is exactly where the new monday.com update comes in.

Integrate your own AI agents directly into monday.com

Many companies have long since built their own agents: research assistants, sales agents, coding agents, or internal tools for specific problems. The challenge until now hasn't been building these agents, but integrating them into the daily workflow of the entire team.

Until now, external agents usually ran in isolation: they helped the person who created them, while the rest of the team continued working elsewhere.

With the monday update, agents from Claude, OpenAI, Azure AI Foundry, Eleven Labs, Microsoft Copilot Studio, Mistral AI, Cursor, Bedrock, and other platforms can be integrated directly into monday.com. These external agents then appear as their own identity in the workspace, alongside monday agents and human team members. You can mention them in conversations, assign them tasks, and collaborate with them just like any other team member—complete with defined permissions and full transparency regarding what the agents are doing.

‍

‍

Work directly in monday.com from Claude, Cursor, ChatGPT, and Slack

Work rarely begins and ends in a single application. Throughout the day, teams constantly switch between tools before the results finally land back in monday.com.

Through monday.com's MCP integration, you can now update boards, assign tasks, and retrieve information without leaving Cursor, Claude, Microsoft Copilot, or ChatGPT.

An example

You are developing a campaign strategy in Claude and want to include data from a previous campaign. Instead of opening monday.com, searching for the right board, switching back to Claude, and manually updating the strategy, you ask Claude directly for the results from the relevant board—and get the information without interrupting your workflow.

Or: A project is being discussed in Slack, and a team member reports that a deadline cannot be met. With the new Slack integration, the project timeline can be adjusted directly from Slack, without switching tools.

‍

‍

Connect your entire AI stack via MCP and open APIs

Every company has a different AI stack. Some teams rely on off-the-shelf AI tools, while others build their own agents and workflows. With MCP support and open APIs, development teams can securely connect AI tools and agents to monday.com—using the same permissions and context as every other user.

New MCP workflow blocks also allow you to build connected automations that involve people, agents, and existing business tools alike. For those who want full control, monday.com’s open APIs allow for building independent, custom integrations that go beyond MCP.

Does this work with Gemini?

In this update, monday.com explicitly mentions Claude, Cursor, ChatGPT, and Microsoft Copilot as MCP tools. Google Gemini is not listed by name, but it has also natively supported MCP since March 2026—so, technically, there is nothing standing in the way of a connection, even if official confirmation from monday.com is still pending.

Why this matters for SMEs

As a monday.com Platinum Partner, we at Blinno regularly see that the biggest leverage in AI projects isn't the individual tool, but whether information actually reaches the places where work happens. A research agent that provides great answers but remains isolated in a chat window offers little value to the team as a whole. This is exactly what this update addresses: agents and AI tools become part of the shared work context instead of forming an additional silo.

For SMEs already using monday.com or planning to get started, it is worth taking a look at your existing AI investments: Which agents or automations are currently running in isolation, and where would connecting them to monday.com immediately remove friction? As with any automation, it remains important to clearly define permissions and data access for agents, especially in the context of GDPR requirements for Swiss companies.

Conclusion

The new AI Work Platform updates from monday.com connect monday agents, external agents, and the AI tools teams are already using into a unified system. Instead of isolated AI successes for individuals, the context becomes usable for the entire team—regardless of whether the work starts in monday.com, Claude, Cursor, ChatGPT, or Slack.