# langchain4j-demo

Companion code for the ankurm.com LangChain4j posts. Every number and transcript quoted in a post
comes from a file in a module's `output/` folder, and the tests that write those files also assert
the same facts, so the build fails when a claim stops being true.

| Module | Post | What it shows |
|---|---|---|
| `ai-services` | LangChain4j AI Services with Spring Boot 4: Declarative LLM Interfaces | prompt templates, chat memory, tools, RAG, and the `@MemoryId` trap |
| `spring-boot-ai-service` | same post | `@AiService` on Spring Boot 4.1.1 with the 1.22.0-beta32 starter |
| `agentic` | Agentic Workflows with LangChain4j: Sequential, Parallel, Loop and Supervisor Agents | the five workflow builders, state passing by key, supervisor JSON, error recovery |

## Versions (verified on Maven Central, 2026-10-11)

| Component | Version | Note |
|---|---|---|
| LangChain4j (`langchain4j`, `langchain4j-core`, BOM) | 1.22.0 | GA |
| LangChain4j beta modules (starters, embeddings, agentic) | 1.22.0-beta32 | beta only, no GA exists; the starter is built against Spring Boot 3.5.13 |
| Spring Boot (`spring-boot-ai-service`) | 4.1.1 | GA |
| JDK | 25 LTS | |
| JUnit | 6.1.3 | |

## Run

```bash
export JAVA_HOME=/path/to/jdk-25
mvn test                                       # everything
mvn test -pl agentic -am                        # one module (-am builds the parent and ai-services too)
```

No API key is needed. The tests use `ScriptedChatModel`, a deterministic stand-in for an LLM, so
they can assert exactly which messages LangChain4j sent. Replace it with `OpenAiChatModel` or any
other `ChatModel` and the AI Service interfaces do not change.

Each module writes its transcripts to `<module>/output/` when you run the tests.