Releases: http4k/http4k
Releases · http4k/http4k
6.15.1.0
Changelog:
- http4k-*: Upgrade versions
- http4k-connect-amazon-instancemetadata: [Fix] Http InstanceMetadataService will now refresh tokens properly. H/T @oharaandrew314
6.15.0.1
Changelog:
- http4k-ai-mcp-sdk: [Fix]
resource
parameter in OAuth corrected (was "resourceUri")
6.15.0.0
Changelog:
- http4k-*: Upgrade versions
- http4k-ai-mcp-sdk: [Breaking] Support for
2025-06-18
spec. Changes to require OAuth resource indicators now require the MCP server URL to be used (to prevent token attack) so if you are using OAuth then you will need to specify it at creation time - http4k-connect-amazon-instancemetadata: [Unlikely Break] Support IMDSv2. H/T @oharaandrew314
6.14.0.0
Changelog:
- http4k-*: Upgrade some dependency versions
- http4k-connect-langchain: [Breaking!] This is now
http4k-ai-langchain4j
. The packages have been moved toorg.http4k.ai.langchain4j
- you can search and replace the imports to fix the breaks. - http4k-ai-mcp-sdk: Add new draft spec features to model.
- http4k-ai-llm-anthropic: [New module!] Universal LLM support for OpenAI.
- http4k-ai-llm-azure: [New module!] Universal LLM support for Azure and Azure GitHubModels.
- http4k-ai-llm-gemini: [New module!] Universal LLM support for Google Gemini models.
- http4k-ai-llm-github: [New module!] Universal LLM support for GitHubModels through the GitHub API.
- http4k-ai-llm-openai: [New module!] Universal LLM support for OpenAI and compatible LLM APis, including Gemini, Azure, GitHubModels
6.13.0.0
Changelog:
- http4k-*: Upgrade some dependency versions
- http4k-server-undertow: [Fix #1365] Gracefully handle unrecognized/invalid HTTP verbs
- http4k-mcp-core: [Breaking] Rename Reference.Resource to Reference.ResourceTemplate.
- http4k-connect-ai-langchain: [Breaking] Rename some classes for consistency. Easy fix :)
6.12.0.0
Changelog:
- http4k-*: Repackaging modules into new http4k ecosystem
http4k-ai
- http4k-ai-llm: [New module!] Core abstractions for a unified LLM API, providing a functional interface for working with LLMs in true http4k style. :)
- http4k-mcp-core: [Breaking] Package migration
org.http4k.mcp
toorg.http4k.ai.mcp
. Fix: Simple replacement will fix imports. - http4k-mcp-core: [Deprecation] Replaced with
http4k-ai-mcp-core
- http4k-mcp-client: [Breaking] Package migration
org.http4k.mcp
toorg.http4k.ai.mcp
. Fix: Simple replacement will fix imports. - http4k-mcp-client: [Deprecation] Replaced with
http4k-ai-mcp-client
- http4k-mcp-sdk: [Breaking] Package migration
org.http4k.mcp
toorg.http4k.ai.mcp
. Fix: Simple replacement will fix imports. - http4k-mcp-sdk: [Deprecation] Replaced with
http4k-ai-mcp-sdk
6.11.1.0
Changelog:
6.11.0.0
Changelog:
- http4k-ai-core: [New module!] Core module for common AI types
- http4k-connect-ai-core: [Breaking] This module has been renamed to
http4k-ai-core
. - http4k-connect*: [Deprecation] Repackaging of core types into
org.http4k.ai
package. Change the offending package names to fix. - http4k-mcp-*: [Deprecation] Repackaging of core types into
org.http4k.ai
package. Change the offending package names to fix.SystemMessage
is nowPrompt
6.10.2.0
Changelog:
- http4k-*: Upgrade some dependency versions.
- http4k-mcp-sdk: Add typesafe lens access to Elicitations.
6.10.1.0
Changelog:
- http4k-*: Upgrade some dependency versions.
- http4k-mcp-sdk: Support for
Elicitations
capability, allowing the server to request User submitted information from the client. This is in the draft spec. - http4k-mcp-client: Support for
Elicitations
capability, allowing the server to request User submitted information from the client. This is in the draft spec.