Link Dump #259

What is the best day for reading? Every day.

1. Software Architecture 
   1. Article Series: Agentic AI Architecture
      The articles, written by industry experts, cover various elements and aspects of agentic AI architecture. They aim to present the latest trends and developments shaping the new type of architecture as it enters the mainstream.
   2. Ontology-Driven Observability: Building an End-to-End Knowledge Graph at Netflix Scale #PickOfTheWeek
      Netflix tackles the challenge of root-cause analysis across massive distributed systems. With millions of concurrent streams and billions of daily requests, cross-domain incidents traditionally required engineers to manually connect fragmented telemetry tools. To bridge these silos, Netflix engineered a machine-readable Observability Ontology. 
      
2. Software Development
   1. What is an AI agent harness? A practical guide
      An AI agent harness provides the surrounding software that helps it get work done. For developers and AI practitioners, this guide explains how harnesses work, the role of context, tools, state and feedback, and what to consider when building one.
   2. SKILL.md Guide: Write a Good Skill for Coding Agents
      A SKILL.md file holds on-demand instructions for a coding agent. This guide covers the format, the description that triggers it, a validator and two tests.
   3. Unobvious consequences of moving the bottleneck #PickOfTheWeek
      The author examines how shifting the primary bottleneck of software development from code generation to reviewing evidence, CI/CD throughput, and intent management fundamentally reshapes the software development lifecycle (SDLC). When AI agents generate code at virtually zero marginal cost, traditional team structures, governance models, and developer skill sets must undergo radical transformation.
   4. Your Project's Constraints Are Upside Down #PickOfTheWeek
      Discover how to untangle real project constraints from organizational habits. In this article, the author explains how tracing rules back to named individual owners and scrutinizing legacy requirements before applying AI automation prevents costly project scope bloat.
   5. Cheaper code won't kill software development
      The author argues that as AI drastically lowers the marginal cost of producing code, software will not disappear; instead, lower costs will dramatically expand software consumption into niche domains, shift operational bottlenecks toward validation and governance, and fundamentally reshape developer career paths and system architecture.
      
3. Leadership
   1. When the Problem Doesn’t Happen
      When a project succeeds or a difficult meeting runs smoothly, most leaders simply celebrate and move on to the next fire. Discover why studying "successful exceptions" is the fastest way to improve team performance. In this article, learn how asking four diagnostic questions turns accidental wins into repeatable, systematic practices.