storytold/lightcraft

LightCraft: An open-source non-destructive photo editor written by scratch in pure Rust

What it solves

LightCraft is an open-source, clean-room reimplementation of Adobe Lightroom, providing a professional-grade photo library and raw development tool. It allows photographers to manage large libraries of images and perform non-destructive editing on RAW files and JPEGs across macOS, Windows, Linux, and the web.

How it works

Built in pure Rust, the application uses a scene-referred, wide-gamut 32-bit float pipeline for high-fidelity color science. It leverages GPU acceleration via wgpu compute kernels (Metal, Vulkan, DX12) for near-instant rendering of adjustments. The software is designed to be "agent-ready," meaning every UI element and engine command is addressable via a JSON-lines control channel and an MCP server, allowing AI agents to drive the application end-to-end.

Who it’s for

Professional and enthusiast photographers who want a high-performance, local-first, and open-source alternative to proprietary photo editing software, as well as developers building AI agents for image processing.

Highlights

  • Non-destructive RAW development: Supports a wide array of camera formats (DNG, CR2, ARW, NEF, etc.) with a high-bit-depth pipeline.
  • Agent-native architecture: Includes an MCP server and scriptable CLI, enabling AI agents to perform tasks like culling, masking, and exporting.
  • High performance: GPU-accelerated rendering allows for slider updates in ~4ms on modern hardware.
  • Local-first privacy: No cloud accounts or subscriptions; the catalog is stored as a human-readable operation log.
  • Advanced Masking: Supports brushes, gradients, and range-based selections (luminance/color) with local adjustments.
  • Cross-platform: Native binaries for all major OSs and a WebAssembly version for the browser.

Written about in

Related

  • Project
    facebookresearch/sam3

    SAM 3 (Segment Anything with Concepts) is Meta’s 848 M‑parameter foundation model that lets you prompt an image or video with free‑form text (or visual exemplars) and receive masks, boxes, and scores for *all* matching objects. It combines a DETR‑style detector and a SAM 2‑style tracker, introduces a presence token for fine‑grained prompt discrimination, and is trained on >4 M auto‑annotated concepts. The repo provides installation steps, example notebooks, and a new SA‑CO benchmark (270 K concepts) for evaluation.

  • Project
    qubvel-org/segmentation_models.pytorch

    A PyTorch library providing a high-level API for image semantic segmentation, featuring 12 architectures and over 800 pretrained encoders.

  • Project
    colmap/colmap

    COLMAP is an open‑source Structure‑from‑Motion and Multi‑View Stereo pipeline that automatically turns collections of photos into 3‑D models. It offers a GUI, command‑line tools, Python bindings, GPU‑accelerated feature extraction, and cross‑platform binaries, making it a go‑to foundation for research and applications in computer vision, robotics, AR/VR, and any workflow that needs camera poses and dense scene geometry.

  • Project
    jacobgil/pytorch-grad-cam

    A PyTorch library for explainable AI in computer vision that provides a wide range of pixel attribution methods to visualize and diagnose model predictions.