1.2.0.CR3

What's Changed

• Ability to run integration tests with Quarkus Platform by @rsvoboda in #1792
• Propagate AWS error body in exceptions from AWS Bedrock by @yurikotikov in #1794
• Make sure that managed QuarkusAiServiceContext is always used by @geoand in #1796

Full Changelog: 1.2.0.CR2...1.2.0.CR3

1.2.0.CR2

What's Changed

• Add the cacheToolList property for MCP clients by @jmartisk in #1743
• Llama and Llama3: Move runtime configuration to RuntimeValue in the Recorder constructor by @gastaldi in #1745
• Patch thought configuration for quarkus-langchain4j-ai-gemini by @hulsbo in #1744
• Remove the Quarkus prefix from the ToC by @gastaldi in #1747
• Only include stopSequences if they are not empty by @geoand in #1753
• Remove quarkus-specific guardrail implementation by @edeandrea in #1754
• Add Content-Type: application/json header to Streamable HTTP MCP client by @geoand in #1755
• [main] Fix MCP integration tests in the Platform CI by @jmartisk in #1756
• Fix non-AI agents support by @geoand in #1763
• Add anthropic thinking config by @ejstuart in #1762
• Fix: ensure USER messages are ordered first for Bedrock (#1764) by @yurikotikov in #1765
• Diff in docs after running the build by @rsvoboda in #1768
• Drop explicit versions of byte-buddy and mockito-core by @rsvoboda in #1769
• Make MCP work with langchain4j agents by @mariofusco in #1758
• Add ChatModelListener support to Bedrock chat models (#1766) by @alexfedorenchik in #1767
• Add codestart artifacts to the generated BOM by @aloubyansky in #1770
• Make sure that the support agents in langchain4j-agentic can be used by @geoand in #1771
• Ability to run integration tests with Quarkus Platform by @rsvoboda in #1777
• Introduce max-completion-tokens config property by @geoand in #1780
• Introduce support for ChatMemoryAccess by @geoand in #1781
• Anthropic handle tool calls by @ejstuart in #1782
• Introduce AiServiceContextFactory by @mariofusco in #1783
• Revert "Ability to run integration tests with Quarkus Platform" by @geoand in #1788

New Contributors

• @hulsbo made their first contribution in #1744
• @ejstuart made their first contribution in #1762
• @yurikotikov made their first contribution in #1765
• @alexfedorenchik made their first contribution in #1767

Full Changelog: 1.2.0.CR1...1.2.0.CR2

1.1.4

What's Changed

• [1.1] Fix MCP integration tests in the Platform CI by @jmartisk in #1750

Full Changelog: 1.1.3...1.1.4

1.2.0.CR1

What's Changed

• Bump to LangChain4j 1.3.0 by @mariofusco in #1679
• Integration tests for exposing resources as tools by @jmartisk in #1691
• Roots integration plus tests by @jmartisk in #1690
• Adapt tests in OpenAI module to run with OpenAI API by @fedinskiy in #1696
• Fix Git text in attributes.adoc by @lordofthejars in #1697
• TLS integration for MCP clients by @jmartisk in #1702
• prevent NPE in case of null parts by @KaiSuchomel in #1704
• Prevent NPE in case of null temperature/topP by @KaiSuchomel in #1714
• Take OLLAMA_MODELS env var into account in the Dev Service by @geoand in #1712
• Introduce support for Audio in AI services by @geoand in #1716
• Feature: Add proxy support for Azure OpenAI and Vertex and Google Gemini by @csotiriou in #1707
• Use the default LangChain4j BedrockStreamingChatModel by @tomas1885 in #1715
• Update AudioUrl.java with correct Javadoc by @lordofthejars in #1718
• Prevent NPE when delta is null in Azure OpenAI streaming by @windschi3f in #1721
• Update quickstart-function-calling.adoc import by @lordofthejars in #1722
• Use a better API to check for fast jar by @gsmet in #1723
• Prevent NPE by @KaiSuchomel in #1724
• Bump mermaid from 11.6.0 to 11.10.1 in /docs by @dependabot[bot] in #1709
• Bump sha.js from 2.4.11 to 2.4.12 in /docs by @dependabot[bot] in #1706
• Handle java time types in json schema generation by @ulisse1996 in #1725
• Bump to LangChain4j 1.4.0 by @mariofusco in #1731
• Use deep link for easy rag docs by @holly-cummins in #1732
• Refactor upstream guardrails by @edeandrea in #1728
• Fix bump to langchain4j 1.4.0 by @mariofusco in #1738
• Introduce the Agentic module by @geoand in #1737
• Send the initialized notification on a worker thread by @jmartisk in #1734
• Bump to LangChain4j Community 1.4.0-beta10 by @geoand in #1739
• Introduce thought configuration to quarkus-langchain4j-ai-gemini by @geoand in #1711

New Contributors

• @windschi3f made their first contribution in #1721
• @ulisse1996 made their first contribution in #1725

Full Changelog: 1.1.1...1.2.0.CR1

1.1.3

What's Changed

• 1.1 backports by @geoand in #1726
• Release 1.1.3 by @geoand in #1727

Full Changelog: 1.1.2...1.1.3

1.1.2

What's Changed

• 1.1 backports1 by @geoand in #1719
• Release 1.1.2 by @geoand in #1720

Full Changelog: 1.1.1...1.1.2

1.1.1

What's Changed

• Set Quarkus LangChain4j version to 1.1.0 in the MCP cli example by @sberyozkin in #1637
• Bump quarkus-neo4j.version from 5.5.0 to 5.6.0 by @dependabot[bot] in #1638
• Change BedrockRecorder to check if the model id contains anthropic in… by @tomas1885 in #1640
• Semantic compression guide by @cescoffier in #1643
• Ollama guide by @cescoffier in #1646
• Streamed response guide by @cescoffier in #1647
• fix: dev-ui being stuck on error by @holomekc in #1650
• Make sure Ollama works in native mode by @geoand in #1645
• Update OidcMcpAuthProvider to ignore requests without active request context by @sberyozkin in #1651
• Update OidcMcpAuthProvider to log how it deals with the access token by @sberyozkin in #1654
• Update Gemini and OpenAi ModelAuthProvider to use ManagedExecutor by @sberyozkin in #1656
• Remove some trash files by @jmartisk in #1659
• Disable the ordered execution of blocking actions by @tomas1885 in #1661
• Support same method name different tool name by @patriot1burke in #1664
• added Record Parser to parse each sse event by @nasonawa in #1657
• Add ability to run integration tests with another version of the extension by @fedinskiy in #1665
• Allow returning a ChatResponse Multi in streaming mode by @tomas1885 in #1662
• Fix for bedrock streaming chat model when there's an error. by @tomas1885 in #1673
• Allow setting maxSequentialToolInvocations on the RegisterAiService annotation by @jmartisk in #1677
• Update MCP integration test dependencies by @jmartisk in #1678
• Ignore static and @JsonIgnore. Look at superclass by @patriot1burke in #1681
• Release 1.1.1 by @geoand in #1682
• Adapt release process to changes in integration tests by @fedinskiy in #1687
• Release 1.1.1 by @geoand in #1688

New Contributors

• @tomas1885 made their first contribution in #1640
• @patriot1burke made their first contribution in #1664
• @fedinskiy made their first contribution in #1665

Full Changelog: 1.1.0...1.1.1

1.1.0

What's Changed

• Implement mcp-sse-client-server locally with ollama and mistral by @diegolovison in #1606
• Add OIDC Client McpClientAuthProvider by @sberyozkin in #1614
• Few-shots, and Image guides by @cescoffier in #1617
• Refer to service account in the command line secure MCP demo by @sberyozkin in #1618
• Bump io.quarkiverse.wiremock:quarkus-wiremock-test from 1.5.0 to 1.5.1 by @dependabot[bot] in #1620
• More secure mcp cmd demo polishing by @sberyozkin in #1621
• Do not use Runtime configuration during deployment by @radcortez in #1623
• Bump quarkus-wiremock to 1.5.1 by @geoand in #1625
• Set configurable timeout for transports by @diegolovison in #1626
• MCP streamablehttp accept header seprated by comma by @nasonawa in #1627
• Use quarkus.langchain4j.timeout as the default MCP timeout by @geoand in #1630
• Check the if id is null or not before starting the SSE channel by @nasonawa in #1631
• Fix milvus dependency version by @iocanel in #1634

New Contributors

• @diegolovison made their first contribution in #1606
• @radcortez made their first contribution in #1623
• @nasonawa made their first contribution in #1627

Full Changelog: 1.1.0.CR2...1.1.0

1.1.0.CR2

What's Changed

• Disable ecosystem CI by @geoand in #1597
• Update status of extensions to "preview" by @geoand in #1598
• Make in-process embedding models respect select model provider by @geoand in #1600
• Dev UI page for MCP clients by @jmartisk in #1601
• Improve Azure streaming test by @geoand in #1604
• Update quickstart-rag.adoc fixing the dependency to langchain4j-embeddings by @AgnolaTommasoDev in #1608

New Contributors

• @AgnolaTommasoDev made their first contribution in #1608

Full Changelog: 1.1.0.CR1...1.1.0.CR2

1.1.0.CR1

What's Changed

• Bump to LangChain4j 1.1.0 by @mariofusco in #1567
• Add integration tests module by @aloubyansky in #1568
• Produce spans for MCP tool calls by @geoand in #1569
• Bump com.google.auth:google-auth-library-oauth2-http from 1.37.0 to 1.37.1 by @dependabot in #1571
• Easy-rag - loading from classpath by @edeandrea in #1570
• Port changes to construct ChatResponse from AzureOpenAiChatModel to AzureOpenAiStreamingChatModel and add test for AzureOpenAiStreamingChatModel. by @dastrobu in #1573
• MCP: Streamable http transport by @jmartisk in #1576
• Revamp documentation by @cescoffier in #1549
• Add Gemini AI Streaming ChatModel by @KaiSuchomel in #1580
• Bump quarkus-neo4j.version from 5.4.0 to 5.5.0 by @dependabot in #1583
• Update to new Ollama api/embed endpoint by @pxnis in #1584
• Rename model-id to model-name in watsonx.ai configuration by @andreadimaio in #1586
• Fix typo in Gemini streaming chat model by @geoand in #1588
• Switch to using DevServicesConfig by @gsmet in #1589
• Bump io.quarkiverse.antora:quarkus-antora from 2.1.0 to 2.2.0 by @dependabot in #1592
• Add MongoDB support for memorystore by @MoheyEl-DinBadr in #1591

New Contributors

• @aloubyansky made their first contribution in #1568
• @KaiSuchomel made their first contribution in #1580
• @pxnis made their first contribution in #1584
• @MoheyEl-DinBadr made their first contribution in #1591

Full Changelog: 1.0.2...1.1.0.CR1