Explore the directory

A harness engineering layer around Codex or Claude Code that turns ad-hoc prompting into a gated workflow of planning, building, self-audit, optional user testing and independent audit before a task can close.

• Coding agents

stable-pretraining is a PyTorch Lightning library that lets researchers swap and compare self-supervised learning objectives, such as SimCLR, DINO, MAE, BYOL, VICReg and LeJEPA, while keeping the same training loop, evaluation and checkpointing.

• Models

QuantumByte is an open-source app-builder that generates a working application from a plain-language product intent, then continuously re-verifies each agent turn against explicit, user-locked business requirements before calling the app ready.

• Agent frameworks

A static analysis tool that checks source code against more than 1,000 bug patterns across many languages, with machine-readable output designed to wire directly into an AI coding agent's own quality checks.

• Code quality

A native macOS terminal that runs sessions in a background daemon so they survive quitting the app, with GPU rendering and a scriptable CLI, and that watches coding agents to flag when one needs approval.

• Coding agents
• Terminals

An MCP server that gives an AI assistant tools to search documents, manage indices, aliases and data streams, and inspect cluster health in Elasticsearch or OpenSearch.

• MCP servers
• Databases

mini-verl adapts the verl reinforcement-learning framework to run PPO, GRPO and on-policy distillation training on a single consumer NVIDIA GPU, reusing existing verl configs and Hydra overrides.

• Models

A native macOS terminal for agent-driven development, built on Ghostty, with an agent-aware sidebar that shows which of many supported coding agents is working, waiting or needs approval.

• Coding agents
• Terminals

A command-line tool that reviews GGUF model files and estimates their memory usage and maximum tokens per second, reading only the file's metadata so nothing has to be downloaded first.

• Inference
• CLI tools

PCILeech Firmware Generator extracts a real PCIe device's configuration over VFIO from donor hardware and uses it to generate unique PCILeech FPGA DMA firmware through a 3-stage host-container-host build pipeline.

• Security

A project-level agent that keeps a project's shared context in a local folder instead of inside one coding agent's session, so switching between Claude Code, Codex or Cursor never requires re-explaining goals or unfinished work.

• Coding agents
• Agent memory

An LLM-native molecular representation that writes molecules as an explicit, typed graph of atoms and bonds instead of a linear SMILES string, so a language model can read, write and edit chemical structure directly.

• Libraries

A discontinued event-based framework for building multi-agent systems, where AI agents communicate asynchronously over a websocket and interact with shared objects in a customizable environment.

• Agent frameworks

A local deep-research tool that turns a folder of PDF, Word, PowerPoint, image and text files into a structured Markdown report, using a judgment model to check evidence coverage alongside the generative LLM.

• RAG

A real-time web dashboard for managing multiple pi coding-agent sessions from one browser tab, with live chat mirroring, an integrated terminal, a diff viewer and mobile remote control.

• Coding agents

A lightweight proxy and load balancer for self-hosted LLM infrastructure, routing requests across local and remote inference backends with automatic failover and a unified model catalogue.

• LLM gateways

An MCP server that runs a structured thought through six specialist AI agents in parallel, each examining it from a different cognitive angle, before a synthesis agent produces one integrated answer.

• MCP servers

A framework for long-context LLM reasoning that replaces free-form recursive code generation with a typed functional runtime grounded in lambda calculus, using the model only on small, bounded subproblems.

• Agent frameworks

A multi-agent system that automates end-to-end multimodal machine learning workflows, turning raw data into a trained ML solution with no manual pipeline building, across CLI, Python API, web UI and MCP interfaces.

• Agent frameworks
• Models

Paper Rebuttal Tips is a Chinese-language guide for writing rebuttals to peer-review comments at AI and machine-learning conferences, organized as reviewer-question, weak-answer, recommended-answer and key-point entries.

• Books

An AI coding agent built for game development, understanding game-engine concepts and generating game assets such as images, 3D models, sprite animation, audio and rigging alongside code.

• Coding agents
• Generative media

A graph-first IDE for agentic coding that shows a live, file-and-import graph of a codebase updating in real time as a coding agent edits files, with per-file shadow history to diff and revert.

• Coding agents
• Editors

A self-evolving multi-agent system for autonomous research and engineering, splitting planning, execution and review across separate roles so long-running work continues, pauses and resumes without a human driving every step.

• Agent frameworks

DeepSec is an AI security platform combining Shield, a code auditor that flags hallucinated packages, missing safeguards and other AI-introduced pattern errors, with Spear, an authorized penetration-testing engine with more than 40 skill packs.

• Security