Lumibot unifies backtesting and live trading for stocks, options, crypto, and forex with AI agent runtime using DuckDB, supporting multiple brokers and data sources.
MCO is a Python CLI tool that orchestrates multiple AI coding agents in parallel, aggregating results with deduplication and consensus, enabling multi-agent review workflows from any AI IDE.
MegaTrain enables training 100B+ parameter LLMs on a single GPU by offloading all parameters to CPU RAM and streaming layers to GPU. Supports HuggingFace models and multi-GPU data parallelism without NCCL.
Memcord is a self-hosted MCP server enabling local-first AI memory with slot-based context isolation and multiple summarization backends, designed for privacy and developer ergonomics.
MemKraft is a zero-dependency local-first memory system storing AI agent knowledge as Markdown, featuring bitemporal tracking, hybrid search, and a prompt self-improvement loop.
MR.ScaleMaster fuses scale-ambiguous monocular SLAM trajectories from multiple robots using Sim(3) graph optimization, enabling heterogeneous SLAM frontends and consistent global maps.
MV-SAM3D extends SAM 3D Objects with entropy-based multi-view fusion and optional pose optimization for more stable and consistent 3D object reconstruction across scenes.
NAS3R enables self-supervised 3D geometry and camera parameter estimation without ground-truth data, using Gaussian splatting and a VGGT backbone. It supports multi-view setups and optional pretrained initialization.
NOVA3R implements a non-pixel-aligned visual transformer for amodal 3D reconstruction from unposed multi-view images, recovering occluded geometry with physical plausibility.
OmniStream uses a multi-frame transformer to process continuous video streams with patch-level temporal indexing, supporting downstream vision-language-action tasks.
OpenMythos implements a recurrent-depth transformer that recycles layers via looped blocks, using input injection to prevent signal drift. It scales from 1B to 1T parameters with up to 1M token context.
OpenResearcher is a fully open 30B agentic LLM designed for deep research tasks, featuring a 96K-turn dataset and a self-built retriever over 11B tokens, running on vLLM with 8×A100 GPUs.
Discover ordinary-claude-skills, a local-first collection of 600+ prompt packages that specialize Claude AI with domain skills, integrated via MCP filesystem for lazy loading.
Otakuapuri is a Python Tkinter app combining manga download, reading, and anime streaming with Cloudflare-bypass scraping and multithreaded UI for responsiveness.
paper-console runs a modular FastAPI backend and Vue/Svelte frontend to print curated content on thermal paper via Raspberry Pi Zero 2 W, supporting dev mock mode and hardware mode with GPIO integration.
PAT3D composes a 9-stage pipeline combining LLMs, vision models, 3D asset generators, and physics simulation to produce physically plausible, simulation-ready 3D scenes from text prompts.
Prefab offers a Python DSL for building declarative UI compiled to JSON and rendered via React, designed for MCP apps and AI agent-generated interfaces without frontend JavaScript.
