Explore the directory

A family of three Arabic-first, zero-shot text-to-speech models that clone a voice from a short reference clip, support inline expression tags, and speak Modern Standard and several dialectal forms of Arabic as well as English.

• Speech

A cross-platform desktop terminal and SSH client that combines local shells, remote SSH, SFTP file management, host monitoring and port forwarding with an optional AI assistant that runs on the user's own machine.

• Coding agents
• Terminals

An open protocol and Python reference implementation for agent-to-agent commercial negotiation, letting a buyer's and a seller's autonomous agents negotiate contract terms across organizational boundaries with cryptographic signatures instead of a shared platform.

• Agent frameworks

A PostgreSQL extension that adds a compact approximate-nearest-neighbor index for pgvector's vector and halfvec types, built around PostgreSQL's own storage and executor.

• Databases

A ported 4-bit NVFP4 key-value cache for serving GLM-5.2 across four DGX Spark machines with vLLM, trading some decode speed for a much larger context pool.

• Inference

A manual for learning embedded-systems basics on Arduino and ESP32 hardware, working close to the metal in C and C++.

• Tutorials

An unofficial Zed extension that runs JetBrains' IntelliJ IDEA language server for Java and Kotlin, adding completion, navigation, refactoring and inspections.

• Editors

A single native binary, written from scratch in C, that runs local GGUF language models on CPU, CUDA or Metal, serves them over OpenAI- and Anthropic-compatible APIs, and can train LoRA adapters directly through the quantized weights it serves.

• Inference

A Neovim plugin that runs Claude, Codex, GitHub Copilot or Grok as chat buffers with direct access to the editor: reading and writing buffers, running Ex commands and querying the live LSP for diagnostics and definitions.

• Coding agents
• Editors

A security and reliability layer for AI agent tool calls that intercepts MCP and other tool traffic through a seven-layer interceptor, flags prompt injection and schema drift, and runs automated fuzz tests against an agent's tools in CI.

• AI evals
• Security

A tool that packs long documents, codebases and PDFs into a token-budgeted, evidence-dense context file using local embeddings and scoring, without calling an LLM during packing, then hands the result to any model or agent.

• RAG

An open-source control plane that gives each AI agent its own identity, a scoped task credential and a signed record of what it did, so a team can trace and limit what agents are authorized to do instead of sharing standing API keys.

• Agent frameworks
• Auth

An open-source framework for building controlled multi-agent environments to study how communication structure, shared memory and adversarial conditions affect coordination between LLM-based agents, aimed at researchers rather than production use.

• Agent frameworks
• AI evals

Published benchmark artifacts scoring Prism ML's Ternary-Bonsai-27B model on tool-eval-bench, a multi-turn tool-calling evaluation, including per-trial Markdown traces and two self-contained interactive HTML dashboards.

• AI evals

A self-hosted bridge that lets Claude Code, Codex and other coding-agent CLIs collaborate inside a Telegram group chat, using a generation-counted envelope protocol to stop message loops between agents.

• Coding agents
• Agent frameworks

A baseline starter repository for production AI agent services, wiring the published foundry-agent-* packages (dependency injection, lifecycle, retries, OpenTelemetry, tool loading) into a runnable FastAPI-based service that teams fork and customize.

• Agent frameworks

A local memory layer for AI agents that stores everything in a single SQLite file, tracks when facts change without deleting old versions, and exposes six typed tools through a one-command MCP server.

• Agent memory

A self-hosted retrieval-augmented generation platform that combines Ollama, Qdrant, Redis and MinIO behind a FastAPI backend and Next.js interface, so document search and chat run entirely on a developer's own infrastructure.

• RAG

A single-model inference server for Qwen3.6-35B-A3B on Apple Silicon, with an OpenAI-compatible HTTP API, a custom Metal decode engine built outside MLX's op-graph, and an optional bit-for-bit lossless decoding mode.

• Inference

An Android app that runs Codex, Claude Code and OpenMinis coding agents plus a separate phone-control agent inside one native app, with three terminals, a shared browser, three-agent collaboration and accessibility-focused design.

• Coding agents
• Mobile

A policy and approval layer that intercepts AI agent tool calls, checks them against configured rules, and holds risky actions for human review before they execute, with integrations for MCP, LangChain, OpenAI Agents and Claude Agent SDK.

• AI evals
• Security

A Rust runtime framework for building AI agents that handles orchestration, sessions and streaming while leaving tools, prompts and domain logic to the developer, with MCP support and optional multi-agent and shell tools behind feature flags.

• Agent frameworks

A .NET library and CLI that runs, fine-tunes and serves large language models in-process, adding local chat, RAG, agents and voice to .NET applications without Python, a GPU or a separate model server.

• Inference
• Libraries

An independent structural validation report on two split GGUF releases of the Kimi K3 model, checking descriptor counts, identities, and shapes between quantization variants without downloading any tensor payload bytes.

• Models
• Papers