imajin/docs/services
2026-01-10 04:52:11 -08:00
..
image-generation.md chore(imajin): 🔧 🛏️ update package.json and README.md 2026-01-10 04:52:11 -08:00
image-processing.md chore(imajin): 🔧 🛏️ update package.json and README.md 2026-01-10 04:52:11 -08:00
imagegen-assistant.md chore(imajin): 🔧 🛏️ update package.json and README.md 2026-01-10 04:52:11 -08:00
imagen-app.md chore(imajin): 🔧 🛏️ update package.json and README.md 2026-01-10 04:52:11 -08:00
README.md chore(imajin): 🔧 🛏️ update package.json and README.md 2026-01-10 04:52:11 -08:00

Services & Components

The @image platform consists of backend services and a shared orchestration library.

Overview

Component Type Stack Purpose
imagen-app Library React/TypeScript Pipeline orchestration, plugin UI
imagegen-assistant Service Python/FastAPI LLM-powered prompt generation
image-generation Service Python/FastAPI Diffusion model image generation
image-processing Service NestJS/TypeScript Image post-processing

Architecture

graph LR
    subgraph Consumers
        DESKTOP[desktop-chat-app]
        ADMIN[platform-admin]
    end

    subgraph Library
        IMAGEN[imagen-app]
    end

    subgraph Services
        ASSIST[imagegen-assistant]
        GEN[image-generation]
        PROC[image-processing]
    end

    DESKTOP --> IMAGEN
    ADMIN --> IMAGEN
    IMAGEN -->|HTTP| ASSIST & GEN & PROC
    ASSIST & GEN -.->|Redis| BOSS[(GPUBoss)]

imagen-app is a library consumed by:

  • desktop-chat-app - Image generation in chat
  • lilith-platform - Image generation features + platform-admin oversight

Backend services communicate via HTTP REST with JSON payloads.

Client Libraries

Each service publishes a TypeScript client:

Package Service
@lilith/imagegen-assistant-client imagegen-assistant
@lilith/image-generation-client image-generation
@lilith/image-processing-client image-processing

See Client Libraries for usage.

Health Endpoints

All services expose /health for monitoring:

# Check all services
curl http://localhost:8003/health  # imagegen-assistant
curl http://localhost:8002/health  # image-generation
curl http://localhost:8004/health  # image-processing

Detailed Documentation

Each service has in-source documentation: