Architecture overview

AI Assistant combines the framework-independent madj2k/ai-core package with TYPO3-specific integration code.

madj2k/ai-core
├── Assistant/      Runtime, context, pipelines and prompts
├── Connection/     Provider contracts and clients
└── Indexing/       Documents, adapters, chunking and vector writes

madj2k/ai-assistant
├── Assistant/      TYPO3 profiles, memory and traces
├── Backend/        Configuration and diagnostics
├── Connection/     TYPO3 connection records and registries
├── Indexing/       TYPO3 sources and persisted indexing state
├── Controller/     Frontend and backend controllers
└── Command/        CLI commands

High-level data flow

flowchart LR
    U[User] --> C[ChatController]
    C --> O[Assistant Orchestrator]
    O --> P[Pipeline]
    P --> Q[Query processing]
    P --> R[Retriever]
    R --> V[Vector store]
    P --> A[Answer generation]
    A --> L[AI connection]
    P --> T[Pipeline traces]
    A --> U

Indexing flow

flowchart LR
    S[Source] --> I[Indexer]
    I --> D[IndexableDocument]
    D --> C[TextChunker]
    C --> E[Embedding request]
    E --> AI[AI connection]
    C --> W[VectorDocument]
    W --> VS[Vector store connection]
    I --> State[IndexerSource state]

Registry-based extension points

Most technical extension points are collected by TYPO3's dependency injection container and exposed through registries. This keeps the core pipeline generic: new processors, connectors, indexers and adapters can be added without changing central orchestration code.