Explore the directory

Moss is a semantic search runtime that embeds directly in an application, built for conversational AI agents that need retrieval latency under 10 milliseconds without running a separate vector database.

• RAG
• Databases

Awesome Adaptation of Agentic AI is a curated list of research papers on adaptation strategies for agentic AI systems, accompanying the paper Adaptation of Agentic AI.

• Papers

Rankify is a Python toolkit for retrieval, re-ranking and retrieval-augmented generation, integrating 40 pre-retrieved benchmark datasets, more than 7 retrieval techniques, more than 24 reranking models, and multiple RAG methods.

• RAG

Kody is a personal-assistant platform that stores an AI agent's memory, keys, code and automations in one place, built on Cloudflare Workers and the Model Context Protocol so the same assistant state is portable across MCP hosts.

• MCP servers
• Agent memory

DevoxxGenie is a Java-based agentic plugin for IntelliJ IDEA that connects to local LLM providers such as Ollama, LMStudio and GPT4All, and to cloud providers such as OpenAI, Anthropic and Gemini, to review, test and explain project code.

• Editors

A feed of indicators of compromise, such as malicious URLs, domains, IPs, and file hashes, gathered from posts shared by the infosec community on Twitter/X, published as CSV, JSON, RSS, MISP, and STIX feeds.

• Security

mlx-dspark runs DeepSeek's DSpark and z-lab's DFlash speculative-decoding drafters natively on Apple Silicon via MLX, giving lossless decoding speedups of up to about 4x for supported consumer-size models.

• Inference

Smithy code generators for Rust that produce clients, servers, and the entire AWS SDK for Rust from Smithy API model definitions.

• APIs

VLM2Vec is the code and data interface for the MMEB family of benchmarks for embedding models, with the latest version, MMEB-V3, covering text, image, video, audio, visual document and agent-centric retrieval.

• AI evals

Talos worker is a downloadable GPU client that pairs with a Talos network account and serves open-model inference jobs from that account's local Ollama installation, reporting uptime toward a share of usage revenue.

• Inference

Illustrated-Agent-Skills is the companion resource repository for a Chinese book about using AI agent skills productively, containing the book's appendices, full writing-workflow skills, chapter example code and data, and the skills used to generate the book's own illustrations.

• Agent skills
• Books

A local-first, general-purpose AI agent app for Android, iOS, and macOS, built on the Pi agent framework, with an extension API for adding tools, settings, and subagents to a mobile agent interface.

• Coding agents

Timber is an ahead-of-time compiler that turns trained classical ML models, such as XGBoost, LightGBM, scikit-learn, CatBoost and ONNX models, into self-contained C99 inference code with no runtime dependencies, and serves them over an Ollama-compatible HTTP API.

• Inference

LangChain.dart is an unofficial Dart port of the Python LangChain framework, giving Dart and Flutter apps a unified interface to language model providers plus components for retrieval-augmented generation and agents.

• Agent frameworks

DataBuff is a self-hosted, AI-native application performance monitoring backend built on OpenTelemetry, with multi-agent troubleshooting that lets LLMs query live traces, metrics and topology directly.

• Monitoring

Awesome A2A is a curated list of resources, implementations, tools, servers and clients related to the Agent2Agent (A2A) Protocol, an open protocol for AI agent interoperability.

• Curated lists

UncommonRoute is an automatic LLM router that plugs into Claude Code, Cursor, Codex or the OpenAI SDK, running locally to send each request to a cheaper or stronger model depending on how complex the request is.

• LLM gateways

This is the official code release for a research paper on benchmarking long-form factuality in large language models, including a fact-seeking prompt set and an automatic evaluator for long-form model responses.

• AI evals

The Elements of Agent Style is a set of 21 English writing rules, formatted for AI coding and writing agents to follow while generating text, so their output for technical documents reads like a tech professional wrote it.

• Agent skills

Solo Agent is an open-source, local-first workspace where humans and multiple AI coding agents, such as Claude Code, Codex, OpenCode, Hermes and OpenClaw, collaborate through channels, threaded conversations and task boards.

• Coding agents

A text-based diagram language where node positions are declared relative to each other, sitting between Mermaid-style automatic layout and the manual, absolute placement of tools such as draw.io.

• Design tools

The Hailo Model Zoo provides pre-trained deep learning models plus a full environment for building, quantizing and evaluating them for deployment on Hailo AI accelerator devices.

• Models
• Inference

A single C++ binary offering 119 speech backends, including 62 text-to-speech engines, for automatic speech recognition, text-to-speech, forced alignment, and translation, with no Python dependency at runtime.

• Speech

pilotfish is a multi-model orchestration policy for Claude Code that keeps a frontier model responsible for planning, approval and final judgment while routing bounded execution work, like search, repetitive edits, tests and documentation, to cheaper models.

• Agent skills