Explore the directory

A deterministic video composition library driven by a single JSON template, rendering the same result on Node, in the browser through WebAssembly, or fully on-device in a React Native app, with no server or generative model.

• Libraries

A self-hosted inference stack for a cluster of NVIDIA DGX Spark machines that routes tasks across local and cloud models, then LoRA-trains the local models on its own routing history so the local share of work grows over time.

• Models
• Inference

A research training method and codebase where a student model learns from a teacher's demonstrations while a constraint limits how far it can drift, aimed at fine-tuning a model on a new task without forgetting old ones.

• Models
• Papers

A local-first, end-to-end encrypted mesh protocol that lets AI agents exchange messages directly over a LAN or LoRa radio with no cloud, internet connection or central server required.

• Agent frameworks
• Networking

A native desktop interface for discovering, downloading, configuring and serving local large language models through llama.cpp, without hand-editing command-line arguments.

• Inference

The source for the Haystack project's website, a static Hugo site with the documentation, blog and tutorials for the open-source LLM application framework.

• Agent frameworks
• Documentation tools

The official Pinecone extension for Gemini CLI, adding Agent Skills and MCP tools for managing vector indexes, running semantic search and building document question-answering assistants from the terminal.

• MCP servers
• Databases

A local, on-device MCP server for macOS that captures everything you have seen, read, typed and heard across your desktop through on-screen OCR, so any connected agent can recall your recent activity.

• MCP servers
• Agent memory

An MCP server that has a second model such as OpenAI Codex or Google Gemini review code that Claude Code just wrote, returning structured feedback inline using a subscription you already hold.

• MCP servers
• Code quality

An internal enterprise framework for building AI agents for support and SRE/DevOps teams, with ready-to-use demo agents, database and messaging integrations, and OIDC, RBAC and mTLS security.

• Agent frameworks

An autonomous offensive-security agent that plans an engagement, runs reconnaissance, chains exploits and writes a report end to end from inside Claude Code, Gemini CLI, Codex or Cursor.

• Security

Three locally fine-tuned Russian-language models for a self-hosted RAG stack: a document splitter, a search-query expander and a retrieval embedder, distilled from a teacher model and quantized to GGUF for llama.cpp on AMD hardware.

• RAG
• Models

A terminal tool for exploring, comparing, monitoring and reporting on open-source package health, adoption, dependencies, releases and security advisories across ecosystems such as PyPI, npm, crates.io and Go modules.

• CLI tools
• Code quality

Research code that reproduces the experiments behind the paper 'MLPs are Hebbians: Constructing Efficient Fact-Storing MLPs for Transformers', including Hebbian MLP construction, Transformer integration and fact-editing experiments.

• Models
• Papers

A testing tool for MCP servers that connects over real stdio, SSE or HTTP transport, checks tool responses against YAML-defined assertions, and lints and fuzzes servers for schema and protocol issues.

• MCP servers
• Testing

A reference collection of Windows commands for system administration, security auditing and incident response, grouped into eight categories such as firewall, logs and persistence checks.

• Security
• Tutorials

A type-safe Lightpanda browser SDK for Laravel and plain PHP, with named instance profiles, automatic CLI or Docker runtime resolution, and MCP integration for fetching and crawling pages.

• Libraries
• Browser automation

A Codex plugin that runs an LSP client in the background so Codex gets post-edit diagnostics automatically, plus explicit MCP tools for language-aware operations such as go-to-definition and rename.

• Coding agents
• MCP servers

A Spring Data module for the Reindexer database, adding repositories, object mapping, type-safe query derivation, transactions and namespace references for Spring and Spring Boot applications.

• Databases

A fully autonomous AI assistant runtime written entirely in Rust, distributed as a roughly 3.4 MB static binary with no Node.js runtime required, designed to run on low-power hardware.

• Agent frameworks

A local Rust AI backend that runs as one daemon so a desktop app, CLI, TUI, scripts and MCP clients all share the same memory, tools, providers and scheduler through one HTTP and WebSocket API.

• Agent frameworks

A drop-in MCP server that gives Claude Code, Cursor, Windsurf and other MCP clients persistent cognitive memory, backed by a Rust engine with temporal decay, contradiction detection and a knowledge graph.

• MCP servers
• Agent memory

A modular local AI agent built with Python and Ollama that calls tools, keeps persistent memory in SQLite, searches the web, and retrieves information from documents with a lightweight RAG pipeline.

• Agent frameworks

An open-source AI agent framework where a developer declares a multi-turn conversation goal in YAML and TrueNorth handles extraction, validation, memory, compliance and output generation.

• Agent frameworks