A continual learning infrastructure for self-improving agents that automates the cycle of serving, collecting feedback, and updating model weights or agent harnesses.

An open-source workspace that integrates multiple AI coding agents into a single board, allowing users to assign issues to agents as if they were human teammates.

Claude Code is an agentic terminal-based coding tool that helps developers execute routine tasks, explain code, and manage git workflows using natural language.

A free, open-source AI chat platform that provides access to multiple language models through a single interface with local browser-based storage for privacy.

An enterprise-grade, local-first agent workbench that unifies multiple AI agent engines into a single workspace for coordinating tasks and creating editable content.

A compact biped robot project providing a full pipeline from MuJoCo-based RL training to real-world deployment on a Raspberry Pi Zero 2 W using ONNX policies.

Semantica is an open‑source Python platform that builds a governed knowledge/Context Graph from enterprise data, adds deterministic reasoning, and records every AI decision as a traceable graph node. It offers SHACL/OWL ontology management, provenance (W3C PROV‑O), polyglot RDF/LPG storage, analytics, and integrations with LangChain, CrewAI, etc., targeting regulated domains that need explainable, auditable AI.

A Node.js server that lets AI coding assistants control a live Chrome browser via the Model‑Context‑Protocol, exposing DevTools features (performance tracing, network/debug console, screenshots, Puppeteer automation) as JSON‑based tools.

A comprehensive multi-agent framework and educational resource that teaches how to build, orchestrate, and evaluate AI agent systems from first principles.

n8n is a fair-code platform for building and deploying AI agents and automated workflows, connecting over 1,500 integrations with flexible model support.

Prime Agent is an open‑source AI coding/research assistant that runs a persistent Python REPL powered by a Recursive Language Model (RLM). It stores supplemental prompts, memories, and reusable skills in a durable “Continual Harness” that can be refined incrementally. The tool supports background, daemon‑backed sessions, sub‑agent spawning, direct agent‑to‑agent communication, scheduled heartbeats, persistent goals, and a bounded autonomous mode. Install via a verified shell script, authenticate a provider, and run `prime-agent` in any project directory. Licensed under MIT and accompanied by extensive documentation and an arXiv citation.

DeepTutor is a lifelong personalized tutoring system that uses agentic AI and multi-engine RAG to transform documents and web sources into interactive, guided learning paths.

lieflat‑gongwen is a data‑driven “skill” for Chinese AI agents that lets them generate official government‑style documents (work plans, research reports, party‑building material) by matching statistical style fingerprints extracted from a 1 M‑character corpus. It detects the document type, applies numeric style constraints, builds a skeleton, fills the text with an LLM, and runs a Python self‑check to ensure the output stays within the measured ranges. The repo includes the rule set, example drafts, and a non‑commercial license.

An AI-powered trading agent framework that integrates market data, quantitative analysis tools, and multi-broker execution for automated trading strategies.

A Rust-based vector index with Python bindings that uses the TurboQuant algorithm to provide high-compression, fast SIMD search without requiring a separate training phase.

A marketplace of ready‑made Cursor AI‑assistant plugins that add agent‑driven capabilities—code review, SaaS integrations (Gmail, Salesforce, GitHub, etc.), workflow orchestration, and learning tools—each packaged with a manifest and optional skill definitions.

A next-generation AI memory management system that emulates human-like cognitive processes to provide agents with dynamic, evolving long-term memory.

A native, GPU-accelerated workspace for managing databases, SSH, SFTP, remote desktops, and AI-assisted server operations in one application.

A centralized memory server and management hub that allows AI agents to share and reuse distilled experience, skills, and project knowledge across sessions and frameworks.

A middleware tool that lets users replace the temporal upscalers and frame generation technologies in games that already support DLSS, FSR, or XeSS.

An agentic toolkit that gives AI assistants direct control over iOS, Android, TV, and desktop apps to automate UI interaction, debugging, and profiling.

An unrestricted, open-source AI studio for generating images, videos, and audio using 400+ models without content filters or subscription fees.

A minimal AI software engineering agent that uses only bash commands to solve coding tasks, providing a high-performance, hackable alternative to complex agent frameworks.

An agent memory system that enables AI agents to learn and build persistent, evolving understanding using biomimetic data structures and multi-strategy retrieval.

Supermemory is a hosted‑plus‑self‑hostable memory and context engine for LLM‑based assistants. It extracts facts from chats, emails, PDFs, images, code, etc., builds per‑user profiles (static facts + dynamic recent context), and provides hybrid RAG + memory search through a single API (`add`, `profile`, `search`). Plugins let popular tools (Claude Code, Cursor, Codex) gain persistent memory, while a one‑binary `supermemory local` server lets you run the whole stack offline with local embeddings. The project claims state‑of‑the‑art performance on LongMemEval, LoCoMo and ConvoMem benchmarks (95 % Recall@15 with >99 % context reduction).

A 45M-parameter tool-calling and structured extraction model designed to run on tiny devices with as little as 28MB of RAM.

An open-source AI video editing tool that allows users to edit raw footage via chat by converting video into structured text and on-demand visuals for an LLM to process.

A Codex-native plugin that provides a rigorous framework for academic research, including specialized workflows for systematic reviews, paper drafting, and peer review simulation.

A fast, GPU-rendered, keyboard-driven IDE that integrates a modular AI coding agent and Unix-style composable tools for power users.

An automatic tool for removing censorship from transformer-based language models using directional ablation and parameter optimization to maintain model intelligence.

A next-generation LLM gateway and AI asset management system that centralizes multi-model API management, format conversion, and usage billing.

An open-source, self-hosted AI assistant that integrates with 20+ messaging platforms and runs on your own hardware to ensure data privacy.

A high-throughput library for LLM inference and serving that uses PagedAttention to optimize memory management and increase efficiency.

Gentle‑AI is a Go‑based configurator that enhances existing AI coding agents (Claude Code, Cursor, etc.) with persistent memory, skill libraries, planning (Spec‑Driven Development), security guardrails and optional evidence‑based review (Receipt‑Driven Development). It installs via a simple script, runs an interactive TUI to select agents and components, writes native config files, and provides a health‑check command. The tool does not ship its own model; it merely augments agents you already use.

A high-performance self-hosted photo and video management solution featuring AI-powered facial recognition and CLIP-based search.

A unified API gateway providing direct, proxy-free access to multiple frontier LLMs like GPT, DeepSeek, and Claude using a standardized OpenAI protocol.

A full-stack Generative UI framework that uses a compact, streaming-first language to let LLMs render structured, interactive components with high token efficiency.

A multi-agent harness that wraps terminal-based AI coding CLIs into a coordinated hive mind, managed by a central orchestrator agent and visualized as a virtual office.

An open-source orchestration platform for managing teams of AI agents through org charts, budgets, and goal-aligned task management.

Ollama is a cross‑platform runtime that lets you download and run open‑source large language models locally. It offers a CLI, a REST API (localhost:11434), and official Python/JS SDKs, plus many community integrations (web UIs, LangChain, AutoGPT, etc.). Install with a one‑line script or Docker, then launch models like `ollama run gemma4` or use `ollama launch` to connect to coding assistants or personal bots.

A draggable, animated whale widget for the DeepSeek Harness web UI that displays your DeepSeek API balance, today’s spend (via a zero‑config bookkeeping mode or an optional real‑time token mode), and per‑turn cost, updating automatically and offering size, sound, and language customisation.

A one-click installation kit that automates the deployment of Qwen3.8-27B on consumer NVIDIA GPUs, automatically selecting the best quantization and context size for the available VRAM.

A 17‑lesson tutorial that teaches you how to build the surrounding “harness” code for an LLM‑based coding agent (Claude). It walks through adding tools, permissions, context management, memory, task scheduling, multi‑agent coordination, and goal‑driven termination, all around a simple agent‑loop that lets the model decide when to call tools.

RLinf is an open‑source, PyTorch‑compatible reinforcement‑learning infrastructure for large‑scale embodied and agentic AI. It ties together many simulators and real‑world robot platforms, supports a wide range of accelerators, and provides ready‑made pipelines for fine‑tuning huge vision‑language, diffusion, and MoE models with RL algorithms such as PPO, GRPO, SAC, and newer flow‑transformation methods. The library includes extensive system optimizations (e.g., FUSCO) and dozens of example recipes, making it suitable for researchers and engineers who need a unified, scalable backbone for training and deploying RL agents in both simulation and the real world.

A plugin marketplace for DeepSeek Harness that enables one-click installation, discovery, and management of community plugins and themes.

Dashi PPT Skill is a Node‑based plugin that lets AI agents generate, edit, and export fully‑styled slide decks. It offers 12 themes, 1 020 templates, rich chart/model layouts, in‑browser controls, and one‑click PPTX/PDF export, all running locally under an AGPL‑3.0 license.

Apache Maka is a high-performance agent workspace that maintains a complete, append-only record of all agent actions to ensure task execution is measurable and reproducible.

img2threejs is a Python‑backed skill that lets an LLM turn a single reference image into a procedural Three.js model. It builds a spec, runs a gated vision‑review loop, and emits TypeScript code that recreates the object from primitives and shaders, ready for animation. The system is extensible via domain plugins (e.g., CS2 weapon skins) and includes a live demo gallery.