In this blog, we show how using Spring AI, we can integrate with Anthropic's Native Skills API for Cloud-Based Document Generation and Custom Skills.

Spring AI adds support for Anthropic's Agent Skills — modular capabilities that let Claude generate actual files rather than text descriptions. With Skills enabled, Claude produces real Excel spreadsheets, PowerPoint presentations, Word documents, and PDFs that you can download and use directly.

This post covers how Spring AI integrates with Skills, including both pre-built Skills and Custom Skills for organization-specific document generation.

…

On behalf of the team and everyone who contributed, we are pleased to announce that Spring for Apache Kafka 4.0.0 is now generally available. We extend our gratitude to all contributors who made this release possible.

Release Journey and Highlights

The development of Spring for Apache Kafka 4.0.0 began in March 2025 with the first milestone release and progressed through five milestone releases and one release candidate over an 8-month development cycle before reaching general availability in November 2025.

This release includes new features, bug fixes, improved documentation, and updated…

In our previous blog post about Anthropic prompt caching, we explored how prompt caching dramatically reduces API costs and latency by reusing previously processed prompt content. We introduced Spring AI's five strategic caching patterns for Anthropic Claude models and showed how they automatically handle cache breakpoint placement while respecting the 4-breakpoint limit.

AWS Bedrock brings prompt caching to a broader ecosystem—supporting both Claude models (accessed via Bedrock) and Amazon's own Nova family. If you're considering Bedrock or already using it, you'll find the same Spring AI…

Large language model API costs can accumulate quickly when applications repeatedly send the same prompt content. A typical scenario: you're building a document analyzer that includes a 3,000-token document in every request. Five questions about that document means processing 15,000 tokens of identical content at full price.

Anthropic's prompt caching addresses this by allowing you to reuse previously processed prompt segments. Spring AI provides comprehensive support through strategic caching patterns that handle cache breakpoint placement and management automatically.

In this blog post, we…

On behalf of the team and everyone who contributed, we are pleased to announce that Spring for Apache Kafka 4.0.0-RC1 has been released. We extend our gratitude to all contributors who made this release possible.

This is our first and final release candidate before the 4.0.0 GA release. We encourage the community to test this release candidate and provide feedback on any issues or concerns before we proceed with the general availability release.

What's New in 4.0.0-RC1

Share Consumer Enhancements: Expanded share consumer capabilities with comprehensive acknowledgment support, concurrency configuration, and factory-level container properties. The share consumer implementation now includes proper poison message protection and delivery count tracking. Speaking of Share Consumers, we recently published a dedicated blog…

Continuing our Road to GA series, this week we're exploring Share Groups in Apache Kafka 4.0.0 and their integration in Spring for Apache Kafka 4.0.0 - a feature that fundamentally expands how we can consume messages from Kafka topics.

When we first start working with Kafka, the mental model is straightforward: topics hold messages, consumers read them, and processing happens in order within partitions. This partition-based model has served countless applications well, providing ordered processing with strong guarantees. However, certain use cases involve creating topics with hundreds of…

On behalf of the team and everyone who contributed, we are pleased to announce that Spring for Apache Kafka 4.0.0-M5 has been released. We also released a patch GA version 3.3.10. We extend our gratitude to all contributors who made these releases possible.

This is the final milestone release for the 4.0 series. We plan to release the first release candidate (RC1) next month.

What's New in 4.0.0-M5

Jackson 3 Support: Added comprehensive Jackson 3 support across all serialization and messaging components. The framework automatically detects and prefers Jackson 3 when available, while Jackson…

On behalf of the team and everyone who contributed, we are pleased to announce that Spring for Apache Kafka 4.0.0-M3 has been released.

We extend our gratitude to all contributors who made these releases possible.

4.0.0-M3 provides initial support for Kafka queues (Share consumers) introduced via KIP-932 in Apache Kafka 4.0.0. Keep in mind that in Apache Kafka 4.0.0, queue support is introduced as an early access feature, and as such, the support that Spring for Apache Kafka provides for this should be considered early access and experimental as well.

You can find more details about it in the reference documentation for Spring for Apache Kafka 4.0.0-M3…

On behalf of the team and everyone who contributed, we are pleased to announce that Spring for Apache Kafka 3.3.7 and 3.2.10 have been released and are available from Maven Central.

We extend our gratitude to all contributors who made these releases possible.

With the 3.2.10 release, we have reached the Open Source End of life for the 3.2.x generation of Spring for Apache Kafka. Please see the support timeline for more details. Please consider upgrading to the latest 3.3.x version for ongoing OSS support.

Release Notes

See the release notes for more details on what's included in these releases…

We are pleased to announce the second milestone release of Spring for Apache Kafka 4.0.x generation - 4.0.0-M2, which now supports Apache Kafka 4.0.0.

Key Changes in 4.0.0-M2

• The Kafka client in 4.0.0-M2 has been updated to 4.0.0
• All ZooKeeper related support is now dropped from the 4.0.0-M2, since Apache Kafka 4.0.0 has completely removed ZooKeeper dependency and now exclusively uses KRaft for leadership election and quorum management
• The EmbeddedKafkaBroker utility that Spring for Apache Kafka provides for testing purposes will no longer work with ZooKeeper, as it only supports the KRaft protocol
• We also verified that the next generation consumer rebalance protocol introduced via KIP-848 works seamlessly via Spring for Apache Kafka 4.0.0-M2
…