Tray.ai announced Agent Gateway, a
new capability in the Tray AI Orchestration platform. IT teams use Agent
Gateway to build governed, maintainable MCP (Model Context Protocol) servers and
MCP tools on Tray and publish them via MCP for stable and secure agent use
across the stack.
As enterprises race to extend agents with new capabilities, shadow MCP servers
and tools are being developed in JavaScript, Python and with ad hoc scripts and
services, often without IT visibility or required guardrails. As Gartner noted
in its September 2025 report, Innovation Insight: MCP Gateways, “Enterprises
are left trying to balance the enthusiasm for adopting MCP, at the cost of
governance, against a more risk-averse approach to adoption, at the risk of
missing the innovation boat.”
Tray Agent Gateway provides a managed environment for creating MCP servers and
tools with defined policies, permissions and versioning, so IT can maintain
security and compliance, reduce shadow MCP development and prepare the agent
ecosystem for emerging standards like A2A (Agent-to-Agent).
“Unmanaged MCP use looks a lot like the API sprawl we saw in the early days of
the cloud era,” said Alistair Russell, Co-founder and CTO of Tray.ai. “API
management centralized and enforced governance across APIs including policies,
permissions and versioning. Tray Agent Gateway does the same for MCP. Now,
developers and business teams can quickly build MCP tools using a unified IDE
for transparency, collaboration and maintenance. With Agent Gateway and our
Enterprise Core working together, MCP servers and tools are all governed,
instrumented and versioned for maximum control.”
Tray Agent Gateway delivers governance for evolving AI standards
Agent Gateway provides the interoperability for MCP and other emerging
protocols that IT teams use to bring rigor, governance and observability to the
enterprise agent stack. Teams can define, test, version and document MCP
services, apply scope and guardrails and share them selectively across their
agent landscape.
Three core ways to use MCP are supported by Agent Gateway:
- Build composite MCP tools: Teams can create sophisticated composite tools in
Merlin Agent Builder and publish them as MCP services. These tools can perform
complex tasks or entire processes end to end and connect to and take action
across any system, from apps to other AI services. Tray Guardian is used to
embed guardrails, and users can mitigate unpredictable agent behavior by encoding
business rules for consistent, auditable outcomes.
- Publish connector-backed MCP tools: Tray’s library of 700+ managed connectors
can now be published as MCP tools, instantly giving MCP-enabled agents secure
reach across CRM, ERP, HR, analytics and other apps, all with the governance
enterprises expect.
- Consume external MCP servers: Tray Agents can securely consume external MCP
servers while IT gains centralized visibility, logging and auditability.
“AI agents will only scale when enterprises standardize how capabilities are
added and governed,” said Rich Waldron, Co-founder and CEO of Tray.ai. “Tray
Agent Gateway gives IT a single way to set policy, permissions and versioning
so teams can reuse trusted tools instead of rebuilding them. The result is
lower risk, lower cost and an operating model that can adapt as standards like
MCP and A2A mature.”
Centralized governance and control at enterprise scale
Agent Gateway gives IT centralized command over Tray-built MCP servers and
tools developed across the organization. By defining MCP servers within Tray
Workspaces and Projects, teams can decide which MCP tools are exposed and under
what conditions, so each MCP deployment aligns with enterprise policies and
security scope.
Every MCP tool and execution is instrumented in Tray Insights Hub, logged and
versioned for traceability. This creates a complete audit trail that can be
streamed to observability platforms such as Datadog or Splunk, giving IT clear
oversight without slowing delivery. With clearly defined ownership, version
control and documentation, IT can maintain governance across hundreds of agents
and tools, replacing fragmented, ad hoc development with a single managed
environment.
Together, Tray Agent Hub and Tray Agent Gateway are part of Merlin Agent
Builder, creating a single environment for teams to build, govern and scale
enterprise agents. Agent Hub provides the composable building blocks for agent
creation, while Agent Gateway delivers governance and control over how those
tools are developed, shared and maintained. With both capabilities in one
experience, Merlin Agent Builder is now the one-stop environment for designing,
deploying and managing agents securely at enterprise scale.
Future-proof architecture for multi-agent interoperability
Agent Gateway is built as a multi-protocol capability, supporting MCP today and
engineered for Google’s A2A and future agent-to-agent interoperability
standards. With this architecture, Tray Agents can communicate, delegate and
collaborate with third-party agents while maintaining enterprise consistency
and control.
By unifying governance across protocols, Agent Gateway future-proofs the
enterprise tech stack, giving teams confidence that their AI orchestration
strategy will remain compatible as new standards emerge.
“Agentic AI and MCP represent revolutionary technology, but they’ve also
introduced a new wave of technical sprawl,” said Marcus Dubreuil, Director of
Systems Architecture at a global music retailer. “We can now use Tray.ai to
spin up new MCP tools seamlessly and on a managed path. IT approves what’s
built, enforces policies and permissions, versions it and publishes it for
secure agent use. From a single platform, we get centralized visibility and
auditability across the MCP servers and tools we deploy-reducing risk from
unvetted code without slowing delivery.”
