In this article, I'll debunk what LLM judges are and go through why they are the best for LLM evaluation.

An overview of the architectural shift from Shallow Agents (Agent 1.0) to Deep Agents (Agent 2.0) and how to build complex AI agents that can handle multi-step tasks over extended periods.

🚀 Curated GPT Image 2 prompt library — OpenAI's next-gen image model with pixel-perfect text rendering, cross-image consistency, and commercial-grade illustration. 16 languages. Free & open...

Notes on context engineering and agent harnesses for video libraries: designing the structured representations to make media legible to LLMs. Since we started joining meetings from our computers, video has become the default way that organizations capture what happens at work. We’re at the point now where recording things

Organize your notes as Markdown files. With native relationships, Git, and Claude Code integration. Free forever.

A personal knowledge base that turns freeform notes into a semantically-connected, AI-augmented knowledge graph.

When should agents use direct APIs vs CLIs vs MCP? Plus patterns for building MCP servers, context-efficient clients and pairing MCP with skills. https://t.co/JEogw5vWly

The leaked source revealed how a production coding agent is actually assembled. Here are the design patterns behind it.

In this article, we will look at what the Claude researchers found.

Bridging the gap between organisational data, implicit organisational reasoning, and AI reasoning with Knowledge Graphs purpose-built for AI.

A unified theoretical framework for structured test-time scaling, showing how topology compression, scope isolation, and decoupled verification—a three-layer structural decoupling—bypass the linear collapse of long-horizon reasoning across multi-agent systems, recursive architectures, and coding agents.

Agent’s system prompt is for identity, constraints, and persistent context. • What the agent is and what it should always or never do • How it should generally approach problems (reasoning style) • Persistent

Claude Code is an agentic coding tool that can run shell commands, edit files, and call external services on behalf of the user. This study describes its comprehensive architecture by analyzing...

Learn how to manage context in Claude Code—when to continue, rewind, compact, or clear a session, and how subagents keep parent context clean.

this is a work in progress mental dump on interesting intersections between how we use and design a harness, implications for memory being accumulated over long timescales, and the search bitter lesson we can’t escape this