Explore the directory

An Android client for the self-hosted Hermes Agent, letting a phone or tablet chat with Hermes sessions over local Wi-Fi or a private Tailscale network, with a merged community Remote Gateway transport.

• Agent frameworks

A passive, offline-by-default tool that builds a decision graph linking every git commit to the coding-agent prompt that produced it, with drift detection, a DLP audit of what agents read, and signed provenance attestations.

• AI evals
• Security

An MCP skill that turns video, audio and screen recordings into searchable, timestamped evidence for AI agents, and checks a frozen contract of file digests, JSON values, SQL or HTTP results by a separate process rather than by model opinion.

• Agent skills
• Testing

A Vim plugin that adds Copilot-style AI code completion and chat, backed by a local Ollama model instead of a cloud service, plus a tool-based editing workflow that can apply changes and iterate on compiler diagnostics.

• Editors

A LangGraph-based general-purpose agent, inspired by OpenClaw's design, that logs every decision, verifies dynamic skill calls with a two-phase help-then-run credential, and delegates work to read-only sub-agents.

• Agent frameworks

A pre-execution firewall for AI agents that intercepts every tool call, classifies it against policy in real time, blocks violations and writes a tamper-evident audit trail, with one line of integration and no agent code changes.

• Security

An open, self-hostable access and governance layer that gives every AI agent in an organization the same identity, policy and audit boundary over MCP or HTTP, with signed, independently verifiable erasure.

• MCP servers
• Security

A modular React Native library for building private, local retrieval-augmented-generation apps, with pluggable LLM, embeddings, vector-store and text-splitter components and on-device inference through ExecuTorch.

• RAG
• Mobile

A Model Context Protocol server connecting Claude and ChatGPT to the Intervals.icu endurance-training API, providing authentication and retrieval of activities, events, wellness data, power curves and custom items.

• MCP servers

A free, MIT-licensed AI coding agent, written from scratch as its own kernel rather than a wrapper, with durable project memory, parallel subagents, an inter-agent mailbox and auto-review agents that critique diffs.

• Coding agents

A family of on-device embodied-AI models for robots: a 1.5B generalist manipulation policy and a 0.9B on-device target tracker, both designed to run locally without a cloud connection.

• Models

An MLX port of Moonshot AI's 2.78-trillion-parameter Kimi-K3 multimodal mixture-of-experts model for Apple Silicon, with a streaming weight converter, expert pruning and per-language expert-overlap analysis, though no single Mac can currently run it.

• Models
• Inference

A zero-intrusion mobile control bridge that sends notifications for local AI coding agent sessions — Claude Code, Codex, Copilot, Gemini, Pi and Grok Build — to Telegram, WeChat or Feishu, and lets you approve actions remotely.

• Coding agents

ElixirSense is a library that provides the building blocks for Elixir developer tooling: code completion, documentation lookup, go-to-definition and function-signature information, combining AST inspection with runtime introspection.

• Editors

A from-scratch machine learning framework built in NumPy, paired with hundreds of browser-graded coding questions, for learning how ML and inference internals actually work instead of only calling an existing library.

• Tutorials

An MCP server that gives an AI assistant a searchable local database of a HackerOne hunter's own bounty history plus 3,600+ public disclosed reports, and generates a full attack briefing for a chosen program.

• MCP servers
• Security

An INT4 weight-quantization method for LLMs that uses learned pairwise rotations to suppress weight outliers, closing most of the accuracy gap to FP16 while running close to AWQ speed on NVIDIA GPUs and Apple Silicon.

• Models
• Papers

A unified multi-backend, multi-device benchmarking utility for Transformers, Timm, PEFT, Diffusers and Sentence-Transformers, covering inference and training under Optimum's hardware optimizations and quantization schemes.

• AI evals

A native Swift CLI and local server that turns an Apple Silicon Mac into a private, OpenAI-compatible AI server for Hugging Face MLX models or Apple's on-device Foundation Model.

• Inference

A Python framework for multi-agent systems where agents communicate, delegate tasks and execute work in parallel, styled as a microservices architecture for AI agents with built-in message queues and observability.

• Agent frameworks

A small, open-source Python CLI coding agent designed to work as well with local and small open models as with large hosted ones, with built-in note-taking, goal tracking and a review step.

• Coding agents

A local-first Model Context Protocol server for managing and semantically searching a document knowledge base, with a built-in web dashboard and every tool also reachable as a plain REST API.

• MCP servers
• RAG

A Go client library for the DeepSeek platform, providing streaming chat completions, thinking-mode reasoning, tool calling, fill-in-the-middle completion, structured JSON output, and an Anthropic-compatible endpoint.

• Libraries

A local bridge that lets Claude Code and Codex collaborate as live peers in the same session — reviewing each other's diffs, negotiating a division of labor, and handing off work when one side runs out of quota.

• Coding agents