Adds the anthropic tokenizer that delegates token counting to the
Anthropic endpoint POST /v1/messages/count_tokens. Since the endpoint
only supports counting, tokenize and detokenize operate on a plain
codepoint list instead of real model tokens.
The local anthropic test profile now uses the new tokenizer.
Uses GET /v1/models to list models and read advertised capabilities.
supportsToolCalling and extended-thinking requests honor the listed
capabilities (including tool_use/tool_calling on compatible servers).
Model listing is cached per profile; unreachable endpoints or missing
capabilities skip the checks rather than failing requests.
Adds AnthropicClient with tool use, SSE streaming, and extended
thinking behind an optional AnthropicDialect bean, plus ANTHROPIC
model provider, request-shaping unit tests, and integration tests
against an Anthropic-compatible server.
- Updated ModelProfile to include OAI_RESPONSES in the ModelProvider enum.
- Created CompletionClientTest to validate text and chat completion functionalities.
- Added ResponsesApiClientTest to test Responses API interactions, including model listing and conversation handling.
- Implemented ResponsesApiDialectTestConfiguration for managing reasoning efforts in tests.
- Updated application-test.properties to include a new profile for responses.
- Added local-responses.properties configuration for the Responses API.
- Added OpenRouterModelDiscovery class for managing model listings and metadata resolution.
- Introduced OpenRouterModelEndpointsResponse and OpenRouterModelsResponse classes for handling API responses.
- Created OpenRouterModelQuery class to encapsulate query parameters for model listing.
- Updated ModelProfile to include OPENROUTER as a model provider.
- Enhanced ModelProfileService to handle default values for maxTokens and maxContextLength.
- Added OpenRouterClientTest for testing model discovery and chat completion functionalities.
- Configured application-test.properties and local-openrouter-chatcompletion.properties for OpenRouter integration.
- Added OllamaClient to handle conversations with the Ollama API.
- Introduced OllamaDialect for managing thinking modes.
- Created data classes: OllamaFunctionCall, OllamaListedModel, OllamaMessage, OllamaShowRequest, OllamaShowResponse, OllamaTagsResponse, OllamaToolCall, OllamaToolDefinition, and OllamaToolFunction for structured API communication.
- Developed unit tests for OllamaClient to validate model listing, chat completion, streaming responses, and tool calling capabilities.
- Configured test properties for local Ollama instance.
- Refactored OpenAiProviderDialect to include methods for managing chat token limits and reasoning effort.
- Updated ChatCompletionClient and AbstractCompletionClient to utilize the new dialect methods for request shaping.
- Modified ChatCompletionRequest to replace reasoning properties with a single reasoningEffort field.
- Adjusted tests to reflect changes in request handling and dialect behavior.
- Updated README to clarify the new request shaping capabilities and defaults.
- Added OpenAiProviderDialect to manage request features for OpenAI-compatible providers.
- Refactored ChatCompletionClient and TextCompletionClient to utilize the new dialect for handling reasoning and usage settings.
- Removed reasoningEffort and reasoningSummaryDetails from ModelProfile and related services.
- Updated README to reflect changes in request shaping and dialect usage.
- Added support for tools in ChatCompletionRequest, including tool choice and parallel tool calls.
- Introduced ChatCompletionTool and ChatCompletionFunction classes to define tool specifications.
- Updated ChatMessage to include tool calls and tool call IDs.
- Created ChatToolCall and ChatToolFunction classes for managing tool call details.
- Modified TextCompletionClient to support tool calling.
- Updated Grader and Grader2 classes to handle new tool-related logic.
- Enhanced Agent and AgentTest classes to incorporate tool calling in agent workflows.
- Added tests for tool calling functionality in CachedClientTest and CompletionClientTest.
- Introduced auto-configuration classes for profile, client, agent, and prompt modules.
- Updated README to reflect new modular architecture and configuration properties.
- Added selective enabling/disabling of modules via application properties.
- Refactored existing services and strategies to remove unnecessary Spring annotations.
- Enhanced test configurations to utilize auto-configuration.
- Updated META-INF imports to include new auto-configuration classes.
- Updated model profile to support new cost parameters: costPerCachedInToken and costPerReasoningOutToken.
- Enhanced AgentRequestStatistic to track cached input tokens and reasoning output tokens.
- Modified CompletionStatistics to include cached input tokens and reasoning output tokens.
- Updated AzureAiClient, CachedClient, and other clients to handle new statistics fields.
- Introduced Timings class to capture detailed timing metrics for requests.
- Adjusted Usage class to include input and output token details.
- Updated properties files to include new cost parameters for cached and reasoning tokens.
