MEGA Code builds a self-evolving infrastructure where AI coding agents improve themselves by automatically extracting Skills and Strategies from session logs
oh-my-claudecode is a multi-agent orchestration framework that layers 27 specialized agents and 28 skills on top of Claude Code
The core concepts of subagent architecture made mainstream by the OpenAI GPT 5.4 nano release, and practical strategies for leveraging it — based on Cole Medin's analysis
Improving the DART client with year fallback and PBR calculation, fixing the market scanner pipeline's missing current_price bug, resolving async/sync confusion in FastMCP middleware, and frontend improvements to SignalPanel and DAG visualization
Practical strategies for getting the most out of Claude Code in real-world use, covering context management, workflow patterns, Second Brain setup, Lazy Loading, TDD-based coding, and subagent usage.
A breakdown of the key changes in Claude Code Skills V2. A built-in benchmarking system measures skill effectiveness numerically, and Skill Creator now automates everything from test case generation to iterative improvement. New frontmatter options and improved implicit triggering are also covered.
How to apply harness engineering in practice using Google's free AI development tool Antigravity. Covers Rules hierarchy, Skills setup, MCP configuration, and building a complete SaaS with vibe coding.
Comparing three open-source projects across the spectrum of AI image generation — classic VGG19 style transfer, Stable Diffusion-based style application, and pose-driven video generation
Adding Google OAuth to FastAPI and React, protecting the full API with JWT cookie-based authentication
Infratice is a problem-based learning platform that provides real-world incident scenarios — from Kubernetes ImagePullBackOff to GitHub Actions build failures — as static logs and config files
