Explore the directory

An open-source safety and control layer for AI agents that blocks unsafe tool calls, pauses for human approval on risky actions, enforces per-agent cost budgets, and keeps an audit trail that can survive a crash.

• AI evals
• Security

A single-binary Rust gateway that caches and routes LLM requests from coding tools such as Copilot, Claude, and Cursor across multiple providers, aiming to deflect most redundant traffic before it reaches the cloud.

• LLM gateways

A drop-in ggml backend that offloads llama.cpp and whisper.cpp prefill computation to Rockchip RK3588 and RK3576 NPUs, leaving sequential decode on the CPU where the NPU has no advantage.

• Inference

A C# library that runs local large language models through the same IChatClient interface used for Azure OpenAI or Ollama, powered by ONNX Runtime GenAI, with Blazor components for building local-model chat interfaces.

• Inference

A Chinese research institute's framework for building supervised, reproducible scientific AI agents and multi-agent systems, built on Microsoft AutoGen and organized around a brain, actuators, memory, and sensors architecture.

• Agent frameworks

An automated code-review command-line agent that reviews a change, applies fixes, and runs tests, recording every step as a structured trajectory used to train an open-weight code-review model.

• Models
• Code quality

Open-source middleware that compiles LLM tool-calling schemas into a reusable key-value cache on disk, so repeated tool definitions are not reprocessed on every request, keeping response latency flat regardless of tool count.

• LLM gateways

An open-source gateway that enforces Model Context Protocol tool-call policy inside a hardware trusted execution environment, so a governed agent process cannot alter or bypass its own policy engine.

• MCP servers
• Security

A Rust-based desktop AI assistant and proxy that runs a tool-driven plan, execute, and review workflow across multiple language models, with a centralized MCP hub, sub-agent delegation, and persistent memory and skills.

• Agent frameworks
• MCP servers

A set of patches and benchmarks that make NVFP4-quantized model inference run correctly and at full speed on Blackwell-generation GeForce and RTX PRO 6000 consumer GPUs, whose architecture differs from datacenter Blackwell.

• Inference

A Microsoft sample application and workshop that builds a retrieval-augmented ChatGPT-style assistant in Java using LangChain4j, Quarkus, and Qdrant as the vector database, deployable to Azure or run fully locally with Ollama.

• RAG
• Tutorials

A dataset-generation skill for AI coding assistants that turns topics, URLs, or raw files into supervised fine-tuning and preference-optimization training datasets, with verification, deduplication, and export pipelines.

• Agent skills
• Models

A command-line tool that reverse engineers a codebase into layered documentation, per-file summaries, per-directory overviews, and runtime-specific entry points, so AI coding assistants understand a project from the first message.

• Agent memory
• Documentation tools

A single gateway that unifies REST, A2A, and Google ADK agents behind one API and one MCP endpoint, converting every response into a uniform stream of AG-UI events with role-based access control.

• Agent frameworks
• MCP servers

The companion code repository for the O'Reilly book Vector Databases: A Practical Introduction, with full working applications for its later chapters.

• Databases
• Books

A cognitive memory system for AI agents that projects semantic embeddings into a high-dimensional phase space to give agents persistent, continuous episodic memory at constant cost, instead of growing linearly with context.

• Agent memory

The first open-source implementation of NVIDIA's KVTC algorithm, compressing LLM key-value caches 6 to 9x using PCA decorrelation, adaptive quantization and entropy coding, to fit longer context in the same GPU memory.

• Inference
• Libraries

A Chrome and Edge browser extension that adds a stack-aware pull request list and a fast in-page review modal to GitHub, so reviewers can browse, diff and comment on stacked PRs without leaving the pulls page.

• Git
• Code quality

A curated collection of research papers on GUI agent safety, organized by environment and keyword, covering benchmarks, datasets, attacks, defenses and evaluation frameworks for computer-use and mobile or desktop agents.

• AI evals
• Security
• Papers

A Visual Studio Code extension that adds syntax highlighting, semantic tokens, formatting, code completion and other language features for CMake scripts, built on an ANTLR4-generated parser.

• Editors

A curated, categorized reference of over 225 AI-powered coding tools, covering full IDEs, editor extensions, terminal and autonomous agents, browser-based builders, and code review platforms.

• Curated lists

A collection of .NET source generators for cross-cutting concerns such as structured logging, OpenTelemetry tracing, retry and cache aspects, DTOs and CRUD APIs, generating code at compile time with no reflection.

• Libraries
• Monitoring

An AWS sample application demonstrating serverless semantic search over images and live video using Amazon OpenSearch as a vector database and Amazon Bedrock's Titan multimodal embeddings model.

• RAG
• Tutorials

An offline, account-free desktop API client built with Kotlin and Compose Multiplatform, supporting HTTP requests, streaming, and Model Context Protocol server testing as an alternative to Postman or Insomnia.

• Testing
• APIs