Explore the directory

A self-extending, sandboxed agentic coding harness built from one loop, two tools and a large language model, able to spawn subagents, pursue a high-level goal until a check command passes, and write its own new tools.

• Coding agents

A CLI for planning whether a large language model will fit and run well on a given machine's unified memory or VRAM, before downloading the weights, with dry-run deployment scripts and benchmarking.

• Inference
• CLI tools

An Xcode extension that adds a local proxy for AI code completion and an agent mode to Xcode 26, letting any OpenAI-format model provider power inline suggestions, codebase-aware edits and git commit messages.

• Coding agents
• Editors

A self-evolving, agentic memory layer for AI agents where facts get superseded rather than overwritten, entity profiles rebuild on demand, and recall never calls a language model, so the same question gets the same answer every time.

• Agent memory

A lightweight local launcher and browser control panel for llama.cpp on Windows, macOS and Linux, covering installation, launching, chat, benchmarking and preset management.

• Inference

A Chinese-language, source-grounded research handbook that dissects seven coding-agent products, such as Claude Code, Codex, Gemini CLI and OpenCode, comparing their agent loops, context handling, tools, safety and multi-agent orchestration.

• Books

A vision-language-action foundation model that turns a vision-language backbone into a real-time robot controller for scientific laboratory tasks, trained with the pi0.5 recipe.

• Models

A grep-like command-line tool that filters lines by a plain-language description of what should match, using a small decision model to answer a yes-or-no question per line instead of a regular expression.

• CLI tools

A Jupyter-based tutorial teaching Python code generation for numerical relativity, using the symbolic-math library SymPy to derive and emit optimized C code for solving Einstein's equations.

• Tutorials

A task-aware MCP server that calculates the VRAM and compute a fine-tuning job needs and returns the cheapest and fastest matching GPUs across a dozen cloud providers, instead of comparing prices by hand.

• MCP servers
• Models

A Vercel AI SDK provider for Ollama that implements the official Ollama API with minimal dependencies, working in Node.js, browsers and edge environments.

• Agent frameworks
• Inference

An agent orchestration system that gives AI engineering agents persistent state outside the model, hierarchical memory, a shared communication layer and a scheduler, so work continues across sessions instead of starting over each time.

• Agent frameworks
• Agent memory

A tool that bridges a codebase and large language models, generating and applying code changes across a project end to end and building curated context prompts from selected files for other AI tools.

• Coding agents

An open, MCP- and A2A-compatible protocol and runtime for auditable human-agent work, giving approvals, overrides, handoffs and escalations one shared, hash-chained shape that can be queried instead of grepped from logs.

• Agent frameworks

A collection of Jupyter notebooks covering agentic AI, multi-agent orchestration, world models and domain-specific LLM fine-tuning, published alongside dozens of models on Hugging Face.

• Tutorials

A Ruby client for the Model Context Protocol that lets an application discover and call external tools over stdio, HTTP or Streamable HTTP, with built-in conversion to OpenAI, Anthropic and Google tool formats.

• MCP servers
• Libraries

Research code for a search-based patch generation method for large-language-model automated program repair, using Monte Carlo tree search and beam search to explore candidate fixes.

• Papers

A mobile runtime detection workbench for iOS and Android that runs Frida-based hooks and shell tools inside a host app sandbox, and lets an AI work in the same sessions through MCP.

• Debugging
• Security

A CLI that provisions and operates a complete self-hosted LLM application stack, including models, vector databases, embeddings, caching and observability, on one private Docker network from a single command.

• Inference
• Hosting

A Java framework for building multi-agent systems, unifying agents, models and tools as composable components, ported to the Java ecosystem from the Python OxyGent framework with Spring Boot integration.

• Agent frameworks

An in-memory knowledge graph server for retrieval-augmented generation that stores entities, relationships and document chunks together with their embeddings so related information stays connected during retrieval.

• RAG
• Databases

A self-hosted, mobile-first multi-model AI chat built on the official GitHub Copilot SDK, letting a user switch between Copilot's models such as Claude and GPT mid-conversation and run autonomous coding agents from any device.

• Coding agents
• Inference

An auditable AI agent control plane with evidence-backed memory, a governed tool runtime, checkpoint-based recovery and a companion engineering tutorial, aimed at answering where a fact came from and what an agent did unattended.

• Agent frameworks
• Agent memory

A project that indexes more than 200 million academic publication abstracts from OpenAlex into a semantic search engine, using embeddings and a fast vector index.

• RAG
• Databases