Explore the directory

A Python framework and API for building and serving long-running AI agents, with durable, resumable runs, agent-to-agent communication, and pluggable tools.

• Agent frameworks

A governance layer that checks the evidence behind an AI coding agent's claims, using git hooks and CI to catch skipped reviews, untested code, and leaked secrets.

• Code quality
• CI/CD

A native Rust desktop app that runs a personal AI assistant with conversations, a native tool system, a plugin standard, and reusable agent graphs.

• Agent frameworks

An open-source orchestrator that turns tickets from an issue tracker into parallel coding-agent sessions, running each in its own workspace with automatic retries and cost tracking.

• Coding agents
• Workflow automation

A VS Code extension that plugs a third-party model gateway into GitHub Copilot Chat's model picker, adding dozens of additional language models through a user-supplied key instead of a separate chat interface.

• LLM gateways
• Editors

A Discord bot that connects Ollama-hosted large language models to a Discord server, letting users chat with a locally hosted model and manage it from Discord.

• Inference

A Burp Suite Professional extension with an embedded MCP server that exposes 150 tools for proxy analysis, active scanning, fuzzing, race-condition testing and guided exploitation to an AI agent such as Claude Code.

• MCP servers
• Security

A Swift library for local computer-use on Apple Silicon: it reads a form in a running app or browser through the Accessibility API and types answers into each field using a small on-device model.

• Models
• Libraries

A self-hosted research harness that turns a query into a comprehensive, fully-cited 'dataroom' of sourced files by running a local model through a search-read-write loop for hours on end.

• Agent frameworks
• RAG

A MySQL-compatible database that adds vector search, ML model training, local LLM inference and RAG pipelines as native SQL calls, collapsing a typical AI data stack into one engine.

• RAG
• Databases

A single Claude Code skill that acts as a full team of 21 role agents, building, reviewing and evaluating a product through a pipeline of code-enforced quality gates from one short brief.

• Coding agents
• Agent skills

An R package that lets R run as an MCP server, so tools such as Claude Desktop or Code and VS Code Copilot can run R code in a live session, and as an MCP client, registering third-party MCP servers with ellmer chats.

• MCP servers
• Libraries

A Claude Code-style terminal coding agent built on DeepSeek, aimed at cutting token cost by more than 90 percent while automatically splitting tasks, developing code and running MCP tests, and interoperating with Claude and Codex.

• Coding agents
• Terminals

A production-ready semantic search API built entirely on Cloudflare's free-tier infrastructure, with OpenAPI documentation and separate reader and writer API keys.

• RAG
• APIs

An extension to ONNX with custom operators for arbitrary-precision integer and minifloat quantization, plus Python utilities for executing, transforming and analyzing quantized neural network models.

• Models
• Libraries

A demo search engine over a catalog of startup descriptions that lets a developer switch between semantic, keyword and hybrid retrieval on the same data, built with Qdrant.

• RAG
• Tutorials

The official research codebase for a paper showing that predicting the next latent state, instead of the next token, trains transformers to learn more compact world models.

• Models
• Papers

A minimal, transparent CLI agent framework that combines Claude Code-style code editing with Manus-style OS control, built around a single readable core file with no hidden abstractions.

• Coding agents
• Agent frameworks

A tool that converts Excel workbooks into structured JSON for LLM and RAG pipelines, and lets an AI agent read and patch-edit Excel files through a CLI and an MCP server.

• MCP servers
• Data engineering

A local-first, proof-carrying agent runtime that records a cryptographically verifiable replay of every model call and tool observation, letting a developer fork a past run at any turn to try a different model or prompt.

• Agent frameworks
• AI evals

A cross-platform local AI-agent runtime and MCP gateway exposing more than 270 tools for local files, Git, processes, Windows automation, WSL, browser control and durable goal tracking to ChatGPT, Codex and other MCP clients.

• Agent frameworks
• MCP servers

An open, free Chinese-language knowledge base of practical AI tutorials, tools and workflows for beginners, covering AI coding agents, content creation and efficiency workflows, maintained by working AI practitioners.

• Tutorials
• Curated lists

A harness for agentic data science that runs an analysis multiple independent ways and checks whether the results converge, instead of trusting a single AI-generated model and conclusion.

• AI evals
• Data engineering

A Swift package for running LLMs such as Gemma 4, Qwen3.5 and Liquid AI's LFM2.5 on iPhone using CoreML, targeting Apple's Neural Engine so the GPU stays free for the rest of the app.

• Inference
• Mobile