AI & Agent
4 articles
The Real Showdown After the Source Leak: Claude Code vs OpenClaw, 510K Lines vs 530K Lines
In our [previous article](https://zmead.com/blog/openclaw-vs-claude-code-design-philosophy), we dissected OpenClaw's 450K lines of open-source code alongside the 183 Markdown config files in Claude Code's public repository. We knew at the time that Claude Code's core runtime was closed-source — all we could analyze was the plugin ecosystem's definition layer. On March 31, 2026, security researcher Chaofan Shou discovered a 60MB `cli.js.map` file buried inside Claude Code v2.1.88's npm package — 1,884 TypeScript source files, 512,664 lines of code, fully exposed. This time, we finally get to crack open Claude Code's real engine and do a truly apples-to-apples source code comparison.
WeChat Meets OpenClaw: What One CLI Command Tells Us About Platform Strategy in the AI Agent Era
WeChat's native integration with OpenClaw signals a major shift in platform strategy. Here's why the world's largest messaging app opening up to AI agents matters.
AI Agents for Advertising: A Practical Guide to Creative Asset Generation
How to build an AI agent system using Strands Agents SDK and Amazon Bedrock to automate advertising creative generation — from copywriting to image and video production.
Building a RAG System from Scratch: A Hands-On Guide with LangChain and Elasticsearch
Step-by-step tutorial on building a Retrieval Augmented Generation system using LangChain, Elasticsearch vector search, and AWS Bedrock — from document embedding to context-enhanced generation.