Explore the directory

A Chinese-language knowledge-sharing project covering large-model interview questions, a chronological history of key papers, and practical courses on agents, RAG and fine-tuning.

• Tutorials

A fast, scriptable command-line client for the App Store Connect API, automating iOS, macOS, tvOS and visionOS release workflows such as TestFlight, builds, submissions and signing.

• CLI tools
• CI/CD

A CLI that fine-tunes and post-trains LLMs from a single YAML config and one command, using layer streaming to train an 8B model on a 4 GB laptop GPU.

• Models

A contributor's guide to working on the Astrid polyrepo, covering its public contracts, contribution tiers, RFC process and release workflow.

• Books

A standalone, framework-agnostic command-line tool that keeps a database schema in sync across developers and production servers using plain SQL migrations.

• CLI tools
• Databases

A pure C/C++ implementation, based on ggml, for running diffusion image, image-editing and video generation models such as Stable Diffusion, Flux, Wan and Qwen Image on the CPU or GPU with no external dependencies.

• Inference
• Generative media

An open-source evolutionary coding agent that turns large language models into iterative code optimizers, discovering new algorithms across generations without human guidance.

• Coding agents

A Python framework for adding input/output guards to large language models, detecting and mitigating risks and generating structured data from LLM output.

• AI evals

The canonical reference documentation for the Astrid runtime, covering its kernel, capsule model, host ABI, inter-process bus and security model.

• Books

EleutherAI's library for training large-scale autoregressive language models on GPUs, built on NVIDIA's Megatron and augmented with DeepSpeed and other optimizations.

• Models

A repository of AI-optimized agent skills, following the open Agent Skills format, grounding coding agents in Android development best practices for tasks where they typically fall short.

• Agent skills

An LLM evaluation platform supporting models from OpenAI, Anthropic, Gemini, Qwen, GLM, DeepSeek and others across more than 100 datasets covering knowledge, reasoning, coding, science and safety.

• AI evals

A CLI and browser extension that lets AI agents read, fill and debug pages in a user's real, already logged-in Chrome or Edge browser, without interrupting the user's own work.

• Browser automation

The companion repository for an eight-week Udemy course on mastering LLM engineering, with weekly projects that build progressively deeper hands-on expertise.

• Tutorials

The official Firecrawl MCP server, giving MCP-compatible AI agents web search, scraping, page interaction and deep research over the live web as clean, agent-ready context.

• MCP servers
• Browser automation

An open-source platform for building versioned, reusable agent skills from a visual workflow, deployable as MCP servers, APIs or SDKs to any agent framework.

• Agent frameworks
• Agent skills

An open-source, self-healing LLM gateway that connects agents and apps to 300+ models through one OpenAI-compatible endpoint, with automatic fallback on failed requests.

• LLM gateways

An experimental JavaScript lexer, parser and interpreter written in Rust, supporting more than 90% of the current ECMAScript specification and embeddable as a Rust crate.

• Languages

An experimental Python library that simulates people with distinct personalities using LLMs, for testing products, generating synthetic data, and gathering persona-based feedback.

• Testing

A Kubernetes-native orchestration platform for reliably running ML pipelines, models and AI agents at scale, defined entirely in Python.

• Data engineering
• Containers

A fully native, lightweight macOS launcher offering app launching, a global hotkey palette, window management, clipboard history and support for running existing Raycast extensions.

• Desktop utilities

A desktop app that provides a graphical interface for OpenClaw AI agents, turning command-line AI orchestration into a windowed experience with no terminal required.

• Coding agents

A set of AI skills distilling a professor's ten years of research and reviewing experience into structured research assistance, covering a paper's full lifecycle from idea to submission.

• Agent skills

A developer tool that turns a browser element selection into source-code context an AI coding agent can use, pointing it straight to the component and file behind what was clicked.

• Coding agents