Vaessl/CLAUDE.md

# CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

## Project Overview

Vaessl is an AI-powered integration bridge that accepts user text/image inputs, processes them through an LLM pipeline (via LiteLLM), and exports structured data to management systems (Homebox, WikiJS). The backend uses a provider pattern for extensibility. The frontend has a working connection management dashboard.

## Commands

### Backend (Spring Boot + Gradle, inside `backend/`)

```bash
./gradlew build           # compile and package
./gradlew test            # run all tests
./gradlew test --tests com.vaessl.app.connection.ConnectionServiceTest  # single test class
```

### Frontend (React + Vite, inside `frontend/`)

```bash
npm run dev       # start dev server
npm run build     # TypeScript check + Vite build
npm run lint      # ESLint
npm run test      # Vitest watch mode
npm run test:ui   # Vitest visual dashboard
```

## Environment

Copy `.env.local` (not committed) into `backend/` with:

- `DB_URL`, `DB_TEST_URL`, `DB_USERNAME`, `DB_PASSWORD` — PostgreSQL (test container on port 5434)
- `PG_DRIVER_CLASS_NAME` — PostgreSQL JDBC driver class
- `OPENAI_KEY`, `OPENAI_BASE_URL` — LiteLLM gateway (provider-agnostic, configured for gpt-4o-mini)
- `FRONTEND_LOCAL_URL`, `FRONTEND_PUBLIC_URL` — allowed CORS origins for the backend

Frontend (optional, defaults to `/api`):

- `VITE_API_URL` — backend base URL used by `api/client.ts`

## Architecture

### Backend (`backend/src/main/java/com/vaessl/app/`)

Package-by-feature layout. Server context path is `/api`. Main endpoints:

- `POST /api/login` — authenticates a service, stores connection ID in session
- `GET /api/connections/status` — lists connected services for the current session
- `DELETE /api/connections/{serviceType}` — removes a service from the session; invalidates the session if no connections remain
- `POST /api/search` — paged search against the requested service; returns 401 if no active session

Six packages:

**`shared/`** — cross-cutting types used by more than one feature package

- `ServiceType` (enum): identifies each integrated app (e.g. `HOMEBOX`); used in both `connection/` and `search/`
- `ServiceProvider` (interface): base for `ConnectionProvider` and `SearchProvider`; declares `getServiceType()`
- `Endpoint` (enum): API path constants for all external service calls
- `SessionKeys`: builds session attribute names of the form `{SERVICE_TYPE}_CONNECTION_ID`

**`config/`**

- `CorsConfig`: env-driven allowed origins (`FRONTEND_LOCAL_URL`, `FRONTEND_PUBLIC_URL`); `allowCredentials(true)` is required for session cookies to work cross-origin
- `SessionConfig`: JDBC-backed Spring Session with a persistent cookie (`SameSite=Lax`, `HttpOnly`)

**`connection/`** — connecting to and persisting service credentials

- `ConnectionProvider` interface: extends `ServiceProvider`; each integrated app implements `login()` and credential checking
- `ConnectionService`: auto-discovers providers via Spring injection, dispatches login by `ServiceType`
- `ConnectionController`: stores `{serviceType}_CONNECTION_ID` in `HttpSession` after login; reads session attributes to build status responses
- Entity (`ConnectionEntity`) uses **Single Table Inheritance** — one `connections` table with app-specific nullable columns
- `HomeboxConnectionProvider` / `HomeboxEntity`: Homebox-specific implementation

**`ai/`** — shared AI infrastructure used by multiple features (search, future image analysis)

- `EmbeddingService`: wraps Spring AI's `EmbeddingModel`; reused by any feature that needs to generate vectors
- `VectorStoreConfig`: Spring bean configuration for `PgVectorStore` (pgvector-backed `VectorStore`)
- All classes here are provider-agnostic — the OpenAI starter is pointed at LiteLLM, so the underlying model is configurable without code changes

**`search/`** — querying connected services (keyword and AI)

- `SearchProvider` interface: extends `ServiceProvider`; each integrated app implements `getSearchResults()`
- `SearchService`: maintains two provider maps — keyword providers and AI providers; routes based on `SearchRequest.aiSearch` flag
- `SearchController`: guards with session check before delegating to `SearchService`
- `SearchRequest`: includes `aiSearch: boolean` — when true, routes to AI provider instead of keyword provider
- `PagedSearchResponse`: includes nullable `summary` field — populated only for AI search results; null for keyword search
- `HomeboxSearchProvider`: keyword search via Homebox API; unchanged from original implementation
- `HomeboxAiSearchProvider`: AI search via pgvector similarity; returns ranked items + generated summary
- `HomeboxSyncService`: fetches all Homebox items page by page, embeds them via `EmbeddingService`, stores in `VectorStore`; triggered on connection login (background sync)

**AI search flow:** on login → background sync indexes all Homebox items into pgvector. On AI search → embed query → pgvector similarity search → top N results passed to `ChatClient` for summary generation → return list + summary. Sync is idempotent (delete-then-reindex per connection).

**`exception/`** — `GlobalExceptionHandler` via `@ControllerAdvice`

### Frontend (`frontend/src/`)

React 19 + TypeScript + SCSS, Vite 6 build. Package-by-feature under `components/`.

- `api/client.ts` — typed `apiFetch` wrapper; always sends `credentials: 'include'` for session cookies; base URL from `VITE_API_URL` (defaults to `/api`)
- `api/connections.ts` — connection-specific API calls (`getStatuses`, `login`, `logout`)
- `api/searches.ts` — search API call (`search`)
- `types/connection.ts` — `ServiceType`, `LoginRequest`, `AuthResponse`, `ConnectionStatus`
- `types/search.ts` — `SearchRequest`, `SearchResponse`, `PagedSearchResponse`
- `components/connections/` — `Dashboard`, `ConnectModal`, `ServiceCard` (connection management UI)
- `components/search/SearchModal` — search dialog; posts to `/api/search` and renders paged results
- `components/ui/` — shared UI primitives (`ActionButton`, `Modal`)

### Data & AI

- PostgreSQL + pgvector (semantic search via embeddings); also used as the Spring Session store (JDBC)
- LiteLLM as a unified AI proxy; Spring AI OpenAI starter wired to it — `OPENAI_BASE_URL` points to LiteLLM, not OpenAI directly, keeping the underlying model provider configurable
- `spring-ai-starter-vector-store-pgvector` provides `PgVectorStore`; configured in `ai/VectorStoreConfig`
- Embedding dimensions must stay consistent with the configured LiteLLM embedding model — changing models requires re-syncing all indexed items
- Processing pipeline (Phase 2): stage in DB → LLM inference → refine via UI → export to target app

### Testing Strategy

Integration tests spin up a **mirrored PostgreSQL container** on port 5434 (same schema as production). WireMock mocks external HTTP APIs (Homebox, WikiJS). Do not mock the database in integration tests — the mirrored container strategy exists specifically to catch schema/migration divergence.