Package atessera.sensei.util

Class LlmService

java.lang.Object

atessera.sensei.util.LlmService

public final class LlmService
extends Object

Factory for creating LLM chat models from SenseiConfig.

Supports two distinct model configurations: writerLlm for composing review responses in the sensei persona, and techLlm for technical tasks such as directory selection, scoring, and run command detection. When one of them is not explicitly configured, it falls back to the other. If neither is configured, the service is considered unavailable.

Constructor Summary

Constructors

Constructor	Description
LlmService(SenseiConfig config)

Method Summary

Modifier and Type	Method	Description
boolean	isAvailable()	Checks whether at least one LLM is configured and available.
dev.langchain4j.model.openai.OpenAiChatModel	newTechChatModel()	Creates a chat model for technical tasks: directory selection, scoring, run command detection.
dev.langchain4j.model.openai.OpenAiChatModel	newWriterChatModel()	Creates a chat model for composing review responses (the sensei persona).

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

LlmService

public LlmService(SenseiConfig config)

Method Details

isAvailable

public boolean isAvailable()

Checks whether at least one LLM is configured and available.

Returns:

true when any LLM-based checks can be performed

newWriterChatModel

public dev.langchain4j.model.openai.OpenAiChatModel newWriterChatModel()

Creates a chat model for composing review responses (the sensei persona).

Uses writerLlm from the config. Falls back to techLlm when writerLlm is not configured.

Returns:

a configured OpenAiChatModel

Throws:

IllegalStateException - when neither writer nor tech LLM is configured

newTechChatModel

public dev.langchain4j.model.openai.OpenAiChatModel newTechChatModel()

Creates a chat model for technical tasks: directory selection, scoring, run command detection.

Uses techLlm from the config. Falls back to writerLlm when techLlm is not configured.

Returns:

a configured OpenAiChatModel

Throws:

IllegalStateException - when neither tech nor writer LLM is configured