kadevin/ilab-conjure

What it solves

iLab CONJURE is a local-first AI image generation workbench that provides a unified interface for managing multiple image generation models. It solves the problem of fragmented workflows by combining prompt engineering, reference image management, and task history into a single WebUI, reducing the need to switch between different AI platforms.

How it works

The project acts as a frontend and management layer that connects to various AI models via different authentication modes. It supports OpenAI-compatible APIs for stable, shared use, and an advanced local OAuth mode for personal workflows using Codex/ChatGPT. The system manages a local SQLite database for task history and uses a local file system for storing generated images and reference assets.

Who it’s for

It is designed for AI artists, designers, and power users who require a structured way to organize their image generation tasks, maintain a consistent library of reference images, and export results for professional use.

Highlights

• Multi-Model Support: Unified access to GPT Image and Gemini models.
• Asset Management: Includes a public image library for reusable reference images and a prompt template system.
• Prompting Tools: Features "chips" for quick insertion of images (@), hex colors (#), and prompt fragments (~).
• Task Management: Supports concurrent tasks, a paginated history library with search/filter capabilities, and ZIP exports of results.
• Integrated Editor: A built-in image editor supporting multi-layer composition, partial erasure, and ratio-locked transformations.
• Deployment Options: Available as a standard application (macOS/Windows) or a portable "unzip-and-run" package.
• CLI Access: Provides a command-line interface for automating image generation and editing tasks.