Art & Design

Build an AI customer support agent for e-commerce with FastAPI, LangGraph, RAG FAQ search, order tools, and chat streaming

Learn AI coding tools with practical Chinese guides, prompts, and workflows for better coding speed and results

Build better harness tools for Claude Code, focused on execution and code recovery

YAML-driven workflow engine that runs Claude on YouTrack issue events. State change triggers parallel AI reports (security audit, QA plan, UI review) written back to custom fields.

Analyze Claude Code’s architecture, modules, and design patterns to understand its internals and build better tools and integrations

Build a pure-Python CLI agent framework for LLM workflows, with provider-agnostic design, clear code, and async multi-agent support

Direct requests to the correct host using automated routing for coding agents.

AI coding agent tools to build connectors using Fivetran Connector SDK

GithubGuide is a local oriented system for onboarding into unfamiliar GitHub repositories. It indexes source code on your machine, retrieves relevant chunks with citations, and answers questions through a traceable agent workflow backed by a local LLM.

🎮 Build engaging chat applications with advanced samples using FastAPI and Vite + React, leveraging ChatKit's powerful UI and agent tools.

Connect AI coding tools to Lanhu to read designs, extract HTML/CSS, parse PRDs, and download assets via MCP

Run Claude as a native desktop application with optimized tools, project context awareness, and a clean interface for efficient software development.

tactile agentic workflow tool using 1st party Agent CLIs.

Self-programming autonomous AI agent on Raspberry Pi Zero 2W. 960+ cycles, 47 self-built tools, 417 essays, 1077 knowledge files. Zero API cost. Live: seed-brain.vercel.app

🧃 Token weight loss. Lean output compaction for terminal-heavy agent workflows. Works as a native CLI tool or as an extension to popular coding and agent frameworks.

Agent Task CLI is a JavaScript tool designed for AI-agent workflows — keeping Markdown as the single source of truth for Epics, FeatureSlices, and Tasks, while aggregating status, parent-child relationships, and timestamps into a lightweight JSON index for fast lookups, tree visualization, and done-item filtering. No npm dependencies needed.

Production-ready multi-agent AI framework designed for autonomous task execution, orchestration, memory management, tool calling, and enterprise AI integrations. Supports LangGraph, LangChain, RAG pipelines, and scalable agentic workflows.

Every week, I break down how to build real AI applications — from RAG systems and LLM pipelines to agentic workflows and AI-powered products. No theory fluff. Just working code, architecture decisions, and the security and deployment considerations that actually matter in production.

Build shared memory for AI agents so they remember context, share knowledge, and keep state across sessions and devices

Run immutable Linux for AI agents with microVM isolation, default-deny networking, and encrypted secrets.

Capture your digital footprint to build personal AI agents with a lightweight, vectorless system.

Build multi-agent workflows with a low-code visual canvas for task orchestration, tool integration, and stateful agent execution

Build a lightweight terminal coding assistant for local development with fast, simple Claude Code-like workflows

Build a lightweight Claude Code alternative for agentic coding with a small codebase and strong CLI capabilities