Apigee MCP Reaches General Availability
Google Cloud's Apigee Model Context Protocol integration has reached general availability. The product enables enterprise teams to expose existing REST and gRPC APIs as MCP tools that agentic AI applications can discover and call — without deploying or managing local MCP server infrastructure. Developers provide OpenAPI Specifications for their APIs; Apigee generates managed MCP endpoints with semantic search in the API hub, so agents can discover capabilities by description rather than by exact name.
The practical value for enterprise teams is significant: organizations that have invested heavily in Apigee for API management can now make those APIs available to internal AI agents and LLM applications with access controlled through existing IAM and API key policies. The approach avoids the proliferation of bespoke per-team MCP servers and instead centralizes tool governance at the API gateway layer. This fits the pattern of "enterprise API hub as AI tool registry" that several vendors are now pursuing.
From a security standpoint, Apigee MCP inherits the authentication, rate limiting, and audit trail capabilities already present in the Apigee platform. Agents calling tools through Apigee MCP go through the same policy enforcement path as human-driven API clients, which is important for organizations with compliance requirements around what data AI systems can access. General availability means production SLA commitments apply.
Read more — Google Cloud Blog
Google Cloud Next '26: Agent Infrastructure Takes Center Stage
Google Cloud Next '26 opens April 22–24 in Las Vegas, and the developer narrative centers on what analysts are calling an "operating system for the agentic enterprise" rather than a model release cycle. Google is reportedly reshaping its Cloud Run, BigQuery, and Gemini offerings into persistent, always-on agent runtimes rather than batch-oriented compute.
Key developer announcements at or leading up to the conference include Gemini 3.1 Pro reaching preview availability on Vertex AI and in Gemini API via Google AI Studio — described as a "noticeably smarter, more capable baseline" for complex problem-solving tasks. Cloud Run Worker Pools reached general availability earlier in April, providing always-on compute for pull-based non-HTTP workloads like queue processors and long-running AI inference. The Cloud Run External Metrics Autoscaler (CREMA), built on KEDA, was open-sourced to enable queue-aware autoscaling based on Pub/Sub backlog or Kafka lag.
Featured developer sessions include live coding demos integrating AI tools into legacy systems via agent orchestration, a session on AI agent interoperability standards (covering MCP, A2A, and related protocols) with participants from Atlassian, Datadog, Harness, and LangChain, and a Google Go runtime performance optimization track with OpenTelemetry observability integration patterns. For teams building on Google Cloud, Next '26 represents the clearest signal yet of where Google's developer platform is heading over the next 12–18 months.
Read more — Google Developers Blog
AWS Smithy Java and Kotlin Client Code Generation Reach GA
Amazon Web Services released general availability for both Smithy Java (April 6) and Smithy Kotlin (April 2) client code generation frameworks. The Smithy Java framework enables developers to build type-safe, protocol-agnostic Java clients directly from Smithy service models. It automates serialization, protocol handling, and request/response lifecycle management — eliminating the manual boilerplate that teams previously wrote and maintained for every AWS service client.
Smithy Kotlin provides the same capability for Kotlin-first codebases: teams define their service model in Smithy IDL and the generator produces idiomatic Kotlin clients that stay synchronized with service API changes automatically. Both frameworks are particularly useful for organizations that maintain internal service clients alongside the AWS SDKs, or for teams building on AWS service APIs where the official SDK lags behind new service features.
The GA releases follow an extended preview period and represent a commitment to API stability for teams adopting the frameworks in production. Both are open source under the Apache 2.0 license and are the same code generation toolchain that AWS uses to produce its own official Java and Kotlin SDKs. Teams working with the AWS SDK for Java or Kotlin will find the tooling conventions familiar.
Read more — AWS Developer Tools Blog
Links & Sources
You May Also Enjoy
