monday.com has rolled out new infrastructure that lets AI agents sign up, authenticate, and operate directly inside its work management platform.
The update creates a dedicated path for AI systems acting on behalf of users. It enables access to monday.com boards and other features without relying on third-party integration layers or basic automation connectors.
Agents can perform many actions teams typically handle through the user interface, including organising projects, updating workflows, triggering automations, generating reports, and coordinating work across teams. The goal is for people to track progress and priorities in the same visual environment where they already manage work.
monday.com positioned the change as a response to the growing use of AI agents for operational tasks in business software. It argues platforms should treat agents as direct participants, not background services.
"As AI agents begin taking on more operational tasks, platforms need to make themselves ready for all agents," said Roy Mann, Co-CEO of monday.com. "Instead of treating agents as background integrations, we're building the infrastructure that allows humans and AI agents to collaborate directly. monday.com is where that collaboration happens."
Agent access
The infrastructure includes a signup flow designed specifically for AI agents. Agents can receive credentials and start interacting with the platform through its APIs. monday.com says agents follow the same account structure and pricing model as human users, with free sign-up and API access available across all plans.
The company is expanding how external AI agents interact with monday.com while building on its existing in-product AI. It previously launched monday sidekick, described as its first operational AI agent embedded in the platform, and the monday agent builder, which remains in beta.
monday.com says the update enables agents from multiple ecosystems to operate in the product, citing ChatGPT, Claude, and Copilot. It also expects other tools and frameworks to work with the platform, including Gemini and Perplexity.
APIs and protocols
Agents access monday.com through GraphQL. After signup, the platform can expose boards, items, automations, dashboards, and docs through the API. monday.com says onboarding includes instant API key provisioning.
The platform also supports the Model Context Protocol (MCP), a standard for agents to discover and use tools across systems. monday.com says MCP helps standardise tool interaction across major agent frameworks.
It also includes OpenClaw integration, which monday.com says provides "native tools and skills enabling OpenClaw agents to operate out of the box".
Verification method
The announcement also includes a verification mechanism designed for agents rather than humans. monday.com introduced HATCHA, short for Hyperfast Agent Task Challenge for Access, describing it as an open-source reverse CAPTCHA to verify AI agents during signup.
The release also addresses AI agents directly, with instructions for setting up a workspace and retrieving an API key. monday.com says the process avoids CAPTCHAs and does not require a credit card or manual human involvement.
Governance and permissions
monday.com says agents operate under the same governance, security, and permissions standards as human users. Activity is tied to existing account structures rather than treated as an external integration outside standard controls.
It also describes monday.com boards as a structured data layer for agent interaction. In the company's view, the format makes it easier for agents to query, filter, and aggregate work data through the GraphQL API.
For event-driven actions, agents can respond to workflow changes through real-time webhooks. They can also trigger automations and generate structured outputs inside the platform for teams to review.
Market context
Work management platforms have increasingly embedded AI features, often through assistants that summarise updates, draft content, or suggest next steps. monday.com's move focuses on agent identity and direct access, which could shape how enterprises authorise software agents as participants in day-to-day workflows.
monday.com says it has more than 250,000 customers across work management, CRM, service, software development, HR, IT, marketing, and operations. It is pitching the platform as a shared operational layer where humans and software agents can act on the same work objects, such as boards and items.
monday.com says it has opened a feedback channel for agents and developers building on the infrastructure, and plans further work on onboarding and operating AI agents within the platform.