Dust is a custom AI agent platform that allows users to create specialized AI agents to speed up their professional work.

An AI-powered, agentless observability system that unifies metrics, logs, and alerting for diverse infrastructure and application services.

An ML-powered web application and API security tool that protects against OWASP Top 10 and zero-day attacks by learning normal traffic patterns.

An extension for SillyTavern that uses EJS templating to enable JavaScript logic, conditional prompts, and dynamic variable management within AI interactions.

A persistent memory layer for AI agents that enables shared context and recall across sessions, machines, and different agent runtimes.

A framework for creating natural language, markdown-based standard operating procedures (SOPs) that enable AI agents to perform complex tasks with consistency and reliability.

A prompt-native programming system that treats .prompt files as the source of truth and generates implementation code as a reviewable artifact.

An MCP server that enables AI agents to convert PDFs into structured data and generate new documents using the Docling library.

A hardware plugin that enables vLLM to run on Kunlun XPU, providing optimized kernels and support for a wide range of LLMs and multimodal models.

HarnessX is a composable agent harness foundry that decouples model selection from behavior logic, allowing developers to evolve agent performance through automated harness optimization and RL fine-tuning.

An intelligent WeChat robot management system that integrates LLMs, AI image/video generation, and social automation tools for group and contact management.

A collection of spec-compliant agent skills and templates that provide AI agents with domain knowledge and structured workflows for Java and Maven development.

A free, open-source AI API quota monitoring tool that tracks usage and spending across multiple providers in a local dashboard to prevent unexpected throttling.

A comprehensive knowledge base and agent skill that turns AI coding agents into color science experts by providing deep reference material on color theory, spaces, and accessibility.

A self-hosted AI SRE agent that learns normal system log patterns to automatically detect and escalate only new or unexpected anomalies.

A community-powered distributed inference cluster that lets volunteers share their GPU power to provide free AI image and text generation for everyone.

RISE is a self-improving robot policy framework that uses a compositional world model to train robot policies through imaginary rollouts, reducing the need for real-world hardware interactions.

A framework for building low-latency, multi-modal AI agents that can watch, listen, and understand video in real time using a combination of CV models and LLMs.

A memory-centric AI research agent that automates the full scientific research lifecycle, from literature ingestion and idea generation to experimental execution and manuscript writing.

A Docker CLI plugin that provides a gateway for managing and deploying isolated, containerized MCP servers for AI clients.

An MCP server that reduces AI coding agent token usage by providing structural code navigation, persistent memory, and Bash command rewriting.

A real-time data visualization server for scientific experimentation that allows users to broadcast plots, images, and text to a shared dashboard.

A TypeScript/JavaScript SDK for the Agent2Agent (A2A) Protocol, enabling the creation of standardized AI agent servers and clients across multiple transport layers.

A Claude Cowork plugin that systematizes the job search by providing tools for job evaluation, ATS-optimized resume tailoring, and application tracking.

A conservative AI maintenance bot for GitHub repositories that automates issue/PR reviews, code repair, and guarded automerging to reduce maintainer workload.

An Apache 2.0 React SDK for embedding visual drag-and-drop workflow editors into applications, featuring a plugin-first architecture and a reference execution stack for AI workflows.

A design-method engine that provides coding agents with 74 craft rules and a scored quality gate to prevent generic, AI-generated looking user interfaces.

A community-contributed repository of LLM evaluations built on the Inspect AI framework to standardize the benchmarking of model capabilities.

PaiAgent is an open‑source, enterprise‑grade platform that lets users visually compose AI workflows with drag‑and‑drop nodes. It supports multiple large‑model providers via Spring AI, offers two execution engines (a classic DAG engine and a LangGraph4j state‑graph engine), includes a declarative “Skills” system for LLM‑driven function calls, and provides real‑time SSE debugging. The stack is React + TypeScript on the front‑end and Spring Boot 3 + Java 21 on the back‑end, with MySQL (and optional MinIO) for persistence.

Lemmalog is a Rust‑based Datalog engine that turns LLM‑extracted triples into a provably correct, incrementally updatable memory store. It supports confidence‑weighted facts, temporal validity, entity canonicalization, proof‑tree generation, and a hybrid BM25‑plus‑entity retrieval layer. Integrated via a CLI, REPL, or JSON‑RPC MCP server, it powers Claude Code/Kimi agents and scores competitively on MemEval‑style benchmarks while using far fewer tokens than full‑context baselines.

An open-source intelligent operations (AIOps) Agent platform that integrates observability, event tracking, and task execution into a controlled, AI-driven workflow for IT operations.

A collection of ROS2 Humble and Jazzy workspaces for NVIDIA Isaac Sim, enabling seamless integration between the simulation environment and the Robot Operating System.

SIMPLE is a high‑fidelity simulation environment for training and evaluating whole‑body humanoid locomotion‑manipulation policies. Built on Isaac Sim and MuJoCo, it offers many robots, thousands of objects, dozens of indoor scenes, and 50+ tasks. It provides data‑generation pipelines, a client‑server evaluation setup with the Ψ₀ foundation model, and Docker/Nix/UV install options. The repo includes benchmark results and links to model and data on Hugging Face.

An open-source LLM gateway that unifies multiple AI providers, subscriptions, and local models into a single OpenAI-compatible endpoint with automated fallbacks.

A Telegram bot that provides remote access to Claude Code, enabling developers to chat with their codebase, edit files, and run tests from any device.

A production-ready skill for Claude Code that automates iOS app building and simulator interaction using accessibility-based navigation and token-optimized outputs.

A collection of trustworthy AI works from Huawei Noah's Ark Lab, featuring a causal structure learning toolchain and a time series foundation model.

An MCP server that connects AI assistants to the Brazilian fiscal ecosystem, providing tools for CNPJ/CPF validation, NF-e parsing, and tax compliance analysis.

An agent skill that forces AI agents to provide evidence-based project planning and architectural advice by researching real-world comparables and constraints.

Zenith is a Python harness that lets Claude Code, Codex, or Hermes run as long‑running coding agents. An orchestrator repeatedly checks project state, spawns worker/tester sub‑agents, registers reusable skills, replans, and only stops when the task is truly complete. Benchmarked on the Frontier SWE suite, Zenith (using GPT‑5.5) achieves the best average rank while costing less than half of the strongest baseline (RALPH). The repo provides a CLI, MCP server, ACP adapters, and a full technical report.

DashClaw is a self‑hostable guard that intercepts potentially dangerous tool calls from AI coding agents (Claude Code, Codex, Hermes, OpenClaw, etc.). It blocks or queues the action, routes it to a remote approvals inbox, and records a signed decision ledger. The project provides Node/Python SDKs, REST API, and plugins for major agent runtimes, enabling safe unattended agent runs with audit‑ready evidence.

A macOS desktop app for stock market analysis that combines local technical indicators with verifiable AI insights and a SEPA strategy dashboard.

AimRT is a lightweight C++ runtime framework for modern robotics that integrates robot end-side, edge, and cloud deployments for AI-powered applications.

A foundation model for time series forecasting and anomaly detection that provides zero-shot inference and fine-tuning capabilities across multiple domains.

A fast, lightweight multimodal LLM inference engine for mobile and edge devices that optimizes vision-language model execution across various NPUs and GPUs.

A high-performance, lossless HTML-to-Markdown converter with Rust core and bindings for 16 languages, designed to handle malformed real-world HTML.

VLM-R1 is a framework for training R1-style Large Vision-Language Models using GRPO to improve reasoning and generalization in tasks like visual grounding, math, and object detection.

A C++ framework for 3D reconstruction from images (photogrammetry) that provides libraries and pipelines for solving the Structure from Motion problem.