AI AG AGENTS is marketed as a solution for planning trips, answering business questions, and solving problems of all kinds, but fixing tools and data outside of their chat interfaces. Developers have to integrate different connectors and keep it running, but that’s a quick method that’s very difficult and creates headaches.

Google says it has solved this by launching its own fully managed, remote MCP servers that will make Google services and Clouds – easier agents to plug into.

The move follows the launch of the The latest model of Google 3 modeland the company is looking to pair more robust reasoning with more reliable connections to global tools and data.

“We made it Google Agent-Ready by Design,” Steren Gannini, Director of Product Management at Google Cloud, told techcrunch.

Instead of spending a week or two setting up connectors, developers can simply hit a URL on a managed endpoint, Giannini said.

At launch, Google started MCP servers for maps, Bigquery, Compute Engine, and Kubertes Engine. In practice, this may look like an assistant asking the assistant directly, or an OPS agent interacting with infrastructure services.

In the case of maps, Giannini said, without the MCP, developers would rely on the model’s built-in knowledge. “But by giving your agent (…) a tool like Google Maps MCP Server, then it is based on the actual, planned location of the area or trips,” he added.

While the MCP servers will eventually be offered across all of Google’s tools, they are initially launching under public preview, meaning they’re not yet fully covered by Google Cloud terms of service. Instead, they are offered at no additional cost to business customers who already pay for Google services.

“We expect to bring them to mass availability in the new year,” Giannini said, adding that he expects more MCP servers to trickle in each week.

MCP, which stands for model context protocol, was launched by Anthropic about a year ago as an open source standard to connect AI systems with data and tools. The protocol has been widely adopted around the world saving the world, and anthropic earlier this week donated MCP to a new Linux foundation fund Dedicated to open sourcing and standardization of AI AG AI infrastructure.

“The beauty of MCP is that, because it’s a standard, if Google provides a server, it can connect to any client,” Giannini said. “I’m looking forward to seeing how many more clients come out.”

One can think of MCP clients as AI apps on the other end of the wire that talk to MCP servers and call the tools they offer. For Google, that includes Gemini Cli and AI Studio. Giannini explained that he also tested it with anthropic and Openii’s Chatgpt as clients, and “they just worked.”

Google argues that it is not about agents connecting to its services. The biggest game in business is Apigee, the API Management Productalready used by many companies to issue API keys, set quotas, and monitor traffic.

Giannini says that apeee can “translate” a standard API to an MCP server, transforming product attributes such as a Product Tool into an agent that can be discovered and used. have security and management controls placed above.

In other words, the same API Guardrails companies use for human-built products can also be used for AI agents.

Google’s new MCP servers are protected by an authorization mechanism called Google Cloud Iam, which clearly protects what an agent can do on that server. They are also protected by Google Cloud Model Armor, which Giannini describes as a firewall dedicated to agenic threats that protects against advanced injection. Administrators can also rely on log logging for additional monitoring.

Google plans to expand MCP support beyond the initial set of servers. Over the next few months, the company will roll out support for services in areas such as storage, databases, logging and monitoring, and security.

“We built the plumbing so the developers didn’t have to,” Giannini said.