Art & Design

The ChatBotKit God SDK offers a Go-based platform for effortlessly building conversational AI bots and agents.

A lightweight, extensible agentic workflow system for small-scale local LLMs, built in TypeScript.

DIY AI Companion inspired by Razer Project AVA. 3D hologram-style avatar, voice interaction, long-term memory, agentic tools, personality that evolves. Built with Python, LangGraph/LangChain, local or cloud LLMs (Ollama/Groq/xAI), Gradio UI, Whisper STT, ElevenLabs or local TTS. Includes DIY Pepper's Ghost hologram guide.

Dependency-free MCP server and agent tool audit starter

解析 Claude Code 源码，提供 14 章 84 篇中文深度笔记，帮助理解 AI Agent 内部实现与工程设计

Run Claude-style CLI workflows through a local LiteLLM gateway with OpenAI, Gemini, DeepSeek, and OpenRouter support

Claude Code harness that builds a live AI dungeon RPG from scratch, then governs it as game master — real-time agent competition with evolving rules

Framework-agnostic GitHub template with an 18-agent Claude Code team, an opt-in Developer Learning Mode, and a 3-domain verification layer (research / implementation / design) with primary-source-only citation. Plan Mode default, bilingual docs, CI invariants, default-off guarantee. Part of the ECC ecosystem.

AI-powered PR review agent on AWS Bedrock (Claude). Reviews GitHub and Azure DevOps pull requests for security, code quality, and requirements coverage, with Jira/Azure Boards work item linking. Deployable as AWS Lambda.

Faraday: An Autonomous Web Research Agent (LangGraph/Streamlit). 🕵️♀️ Investigates queries using dynamic tools (Tavily, Google, NewsAPI, etc.), gathers multi-source info, and synthesizes structured reports in a Streamlit UI. Features agentic workflow & source tracking.

Secure agentic tool-use loop build with the Claude API - CCA-F study project

AI Agent Orchestration Platform — multi-agent workflows with visual canvas, multi-LLM support, custom tool building, Telegram integration, and template library.

Automation and wallet intelligence toolkit for the Arc ecosystem. Built for onchain monitoring, UserOps tracking, gas analysis, and future AI-agent workflows.

An open-source agentic workflow builder that generates, executes, debugs, and solidifies workflows through sandbox-driven agents.

A lightweight command-line interface (CLI) starting point for building agent workflows using Autodesk Platform Services (APS) APIs. Designed to be extended (by agents) for specific workflows, it comes bare-bones by design.

Building tools for AI agent workflows

A hands-on GenAI journey exploring LLM fine-tuning, RAG, diffusion models & agentic workflows. Built with PyTorch, LangChain & Hugging Face, this repo documents experiments in prompt engineering, RLHF, and AI app deployment. Less pure theory, more code because the absolute best way to master AI is by building and breaking things.

A collection of lightweight, stdlib-only Python utilities built for AI agents

Hermes agent toolchain — one-click setup scripts for Hermes CLI, providers, Chrome MCP, tmux, Ghostty

Build autonomous, immutable multi-agent workflows in Elixir using pure functional agents and a reliable OTP runtime.

Build AI agents that join, listen, and speak in X/Twitter Spaces with real-time transcription, LLM replies, and voice synthesis

Agent skill for LaTeX-based bilingual PDF translation with page-level visual QA.

AI sales mentor for real estate agents - workflows, listings, scripts, follow-up and operational support

SSOT for shared context used by TRAVLR agents, tools, and services