# Perspectives Published articles for Perspectives. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Knowledge as Code: The Memory File Just Got a Spec DevFeed: [Knowledge as Code: The Memory File Just Got a Spec]() Original publisher: [Read original article]() Author: Engin Diri Published: 2026-07-14T00:00:00Z Content type: opinion Language: en Sources: [Pulumi]() Topics: [Wiki](), [Markdown](), [Retrieval Augmented Generation (RAG)](), [context window]() Tags: [agent](), [ai](), [ai-agents](), [google](), [google-cloud](), [knowledge-base](), [memory](), [perspectives](), [platform-engineering](), [rag](), [schema]() ### AI overview The article argues that agent memory files need a shared format so knowledge bases can be exchanged between people and agents. It describes Google's LLM wiki pattern as an approach using interlinked Markdown pages, summaries, change logs, and entity pages, contrasting it with repeatedly retrieving and re-deriving information from raw documents. ### Source excerpt Five weeks ago I wrote that the least glamorous piece of an agent loop is also the one that decides whether it compounds: memory. A markdown file outside the context window that holds what is done, what is next, and what was learned, because the model forgets all of it between runs. Write the memory file before the loop. What I left open, because there was nothing to point at, was the format. My memory file looked nothing like yours, and neither of our agents could read the other's. Three days after that post went live, Google shipped an answer. The pattern everyone copied Andrej Karpathy published a gist in April he called the LLM wiki, and it collected thousands of stars and forks. It's meant to be pasted straight into a coding agent. Instead of indexing your documents for RAG and re-deriving answers from raw text on every query, the agent builds a wiki and keeps it current: interlinked markdown pages, an index.md with a one-line summary per page, a log.md recording every change, entity pages that grow as sources come in. Drop in a meeting transcript and the agent reads it, updates a dozen existing pages, fixes the cross-references, and appends to the log in one pass. It took off for the same reason wikis usually die. A knowledge base is valuable in exact proportion to the bookkeeping nobody wants to do: summarizing, linking, reconciling contradictions, pruning stale claims. Karpathy's line, which Google now quotes back in its own announcement, is that LLMs "don't get bored, don't forget to update a cross-reference, and can touch 15 files in one pass." The human curates sources and asks questions. The agent does the janitorial work that made every previous wiki rot. A wiki only your agent can read Then everyone built one, and every one is a dialect. Mine links related pages in the frontmatter; yours links them at the bottom of the body. Mine has a tags field; yours calls it categories. None of this matters while the wiki serves one person, because the schema lives ## Loop Engineering for Coding Agents DevFeed: [Loop Engineering for Coding Agents]() Original publisher: [Read original article]() Author: Engin Diri Published: 2026-06-09T00:00:00Z Content type: opinion Language: en Sources: [Pulumi]() Topics: [Loop Engineering](), [Claude Code](), [Artificial Intelligence](), [Tooling]() Tags: [agent](), [ai](), [ai-agents](), [automation](), [claude](), [claude-code](), [perspectives](), [platform-engineering](), [prompting](), [pulumi-neo](), [tooling]() ### AI overview An opinion article argues that work with coding agents is shifting from manually writing successive prompts to designing automated loops that identify work, prompt agents, evaluate results, and select next steps. ### Source excerpt For about two years, the unit of work with a coding agent was the prompt. You wrote a good one, you gave it enough context, you read what came back, and you wrote the next one. The agent was a tool, and you were holding it the entire time, one turn after another. That part is ending. Addy Osmani, a director of AI at Google Cloud, has a name for what replaces it, and I have not stopped thinking about it since: loop engineering. You stop being the person who prompts the agent. You design the loop that prompts it for you. In my phrasing: you stop being the thing that runs, and start designing the thing that runs. The leverage moves up a layer. What I want to do here is take an honest look at the pieces, and at the part nobody automates. The leverage moved up a layer The people building these tools have already made the jump. Peter Steinberger has been posting it as a monthly reminder. Peter Steinberger (@steipete) on X. Boris Cherny, who heads Claude Code at Anthropic, says the same thing about his own job. He does not prompt Claude anymore. He has loops running that prompt Claude and decide what to do next, scanning the issue tracker, the team chat, and the timeline for what to build. "My job is to write loops." A loop is a goal that prompts itself. You set the purpose, and the system keeps iterating until it's met. In practice it finds the work, hands it out, checks the result, writes down what got finished, and decides the next thing, then it pokes the agent instead of you. You build that small system once and let it run. Look closer, and a loop is really two loops nested. The inner one does the work against a spec. The outer one decides what the work should be: it watches an issue tracker, an error feed, a changelog, then writes the next spec and hands it down. Most people are still running that outer loop by hand, in their head, and calling it a backlog. The part that surprised me is that this is barely a tooling problem anymore. A year ago a loop meant a pile of ## Build a Repository Harness for Claude Code Coding Agents DevFeed: [Build a Repository Harness for Claude Code Coding Agents]() Original publisher: [Read original article]() Author: Engin Diri Published: 2026-05-26T00:00:00Z Content type: opinion Language: en Sources: [Pulumi]() Topics: [Claude Code](), [AI-assisted coding](), [Tooling](), [Code](), [Command-line interface](), [MCP]() Tags: [ai](), [ai-agents](), [claude](), [claude-code](), [coding-agents](), [devops](), [extension](), [hooks](), [ls](), [mcp](), [perspectives](), [plugins](), [skills](), [tooling]() ### AI overview The article argues that, in real codebases, the context and tooling surrounding Claude Code can matter more to performance than the underlying model version. It presents a repository-level AI harness built from seven components, including CLAUDE.md, hooks, skills, plugins, LSP, MCP, and subagents, and discusses agentic navigation without a separately maintained codebase index. ### Source excerpt Anthropic shipped a piece earlier this month called How Claude Code Works in Large Codebases. I have not read anything more useful about coding agents this year. The core claim, in their words: "the ecosystem built around the model--the harness--determines how Claude Code performs more than the model alone." In my phrasing: in a real codebase, the model is the smaller variable. The layer of context and tooling you wire around the agent matters more than which version of Sonnet or Opus is behind it. The post stays high-level, which is the right move for a launch piece. What I want to do here is land it. Same seven pieces, but with the wiring you would actually put in a repo, in the order I would put it. How Claude Code navigates without an index Anthropic's writeup says Claude Code works from the live codebase and does not require a codebase index to be built, maintained, or uploaded. The agent navigates the way an engineer would, with grep, find, ls, file reads, and reference-following. Anthropic calls this agentic search, and the upside is obvious: no separate index exists for you to keep fresh. The downside is also obvious. An engineer who has never seen your repo and only has shell tools will flounder if you drop them in the root with no map. That is your agent on day one. Everything that follows is about giving it the map. The AI layer in seven pieces Every codebase used to have two artifacts engineers cared about: the code and the tests. A third exists now. Call it the AI layer, or the harness, or whatever you want. This layer is the set of context and tools you give your coding agent to operate in this specific repo. Anthropic breaks it into seven pieces, and each one solves a different scaling problem. Anthropic gives each piece a role: CLAUDE.md is the foundation, hooks do self-improvement, skills are progressive disclosure, plugins handle distribution, LSP gives navigation, MCP is extension, subagents split exploration from editing. They are not equal in usag ## New Leaders, New Perspectives: Why We Joined Wealthfront Engineering DevFeed: [New Leaders, New Perspectives: Why We Joined Wealthfront Engineering]() Original publisher: [Read original article]() Author: Sanjana Kaundinya Published: 2026-02-12T23:10:24Z Content type: article Language: en Sources: [Wealthfront]() Topics: [engineering-culture](), [DevOps](), [Back end](), [Kafka]() Tags: [backend](), [devops](), [engineering](), [engineering-culture](), [kafka](), [perspectives](), [wealthfront-engineering]() ### AI overview Wealthfront interviews two new engineering managers about joining the company, its engineering culture, and how individual ownership shapes software design and project direction. ### Source excerpt At Wealthfront, our engineering culture is something you experience as soon as you join the organization, and new hires are regularly surprised by how quickly and fully they begin to contribute and drive impact. We sat down with two of our new engineering managers, Melissa and Sanjana, to hear how quickly they were able to... Read more ## Does AI Get Bored? DevFeed: [Does AI Get Bored?]() Original publisher: [Read original article]() Published: 2025-09-27T00:00:00Z Content type: opinion Language: en Sources: [Tim Kellogg]() Topics: [Artificial Intelligence](), [Large Language Model](), [Code](), [GitHub]() Tags: [ai](), [experiment](), [github](), [llms](), [opinions](), [perspectives]() ### AI overview The author experiments with giving AI models extended periods without assigned tasks, sometimes providing tools such as drawing, search, fetching, and time manipulation. They report behaviors they call "collapse" and "meditation," while acknowledging uncertainty about how to interpret the results. ### Source excerpt We always give AI something to do. Chat with us, do tasks for us, answer questions, parse text. What happens when we give an AI nothing to do? I didn't know, so I tried. ## Devconnect Scholars Program - Ethereum Stories from Istanbul and Beyond DevFeed: [Devconnect Scholars Program - Ethereum Stories from Istanbul and Beyond]() Original publisher: [Read original article]() Author: Next Billion Published: 2024-02-29T00:00:00Z Content type: opinion Language: en Sources: [Ethereum Foundation Blog]() Topics: [Ethereum](), [Resilience](), [Protocol (disambiguation)](), [Blockchain]() Tags: [africa](), [backend](), [banking](), [bitcoin](), [blockchain](), [community](), [devconnect](), [diversity](), [ecosystem](), [ethereum](), [geography](), [india](), [innovation](), [iran](), [java](), [network](), [networks](), [next-billion](), [perspectives](), [program](), [resilience]() ### AI overview The article presents stories from participants in the Devconnect Scholars Program, describing how broader geographic and demographic representation can strengthen Ethereum's resilience, security, and community. It profiles scholars from India, Iran, and Kenya and their work in research, education, smart contracts, and open-source projects. ### Source excerpt Ethereum is growing, and diversity of human participation creates resilience throughout the ecosystem. The Devconnect Scholars Program is one small effort that aims to create resilience through community diversity. Better representation across human geography and demographics leads to diverse experiences and new perspectives that help the Ethereum protocol serve the... ## Perspectives on React Native from Three Shopify Developers DevFeed: [Perspectives on React Native from Three Shopify Developers]() Original publisher: [Read original article]() Author: Ash Furrow Published: 2021-07-16T13:00:00Z Content type: article Language: en Sources: [Shopify Engineering](), [Shopify Engineering - Shopify Engineering]() Topics: [React Native](), [Mobile](), [Android](), [iOS](), [Web](), [Development]() Tags: [android](), [developers](), [ios](), [learning](), [mobile](), [native](), [perspectives](), [react-native](), [shopify](), [web]() ### AI overview This article presents perspectives from three Shopify developers on React Native. It explains how React Native lets web developers use the React paradigm to build iOS and Android user interfaces, while helping native developers build interfaces faster and reach a wider audience. The article also discusses collaboration across web, Android, and iOS backgrounds, along with the challenges and experiences of learning React Native. ### Source excerpt We hope that by sharing our experiences learning React Native, we can help soothe your anxieties and empower you to explore this exciting technology. ## Two Perspectives on the End-to-End Principle DevFeed: [Two Perspectives on the End-to-End Principle]() Original publisher: [Read original article]() Author: Nelson Elhage Published: 2017-06-11T20:42:01Z Content type: opinion Language: en Sources: [Nelson Elhage]() Topics: [systems](), [Architecture & Design](), [Complex Systems](), [Software Engineering]() Tags: [argument](), [complex-systems](), [perspectives](), [software-design](), [systems-engineering]() ### AI overview The article reflects on the end-to-end principle as a general systems-design heuristic. It explains that functions such as correctness may be better handled at the ends of a system rather than at every lower-level interface, and questions whether complex systems can be made correct simply by composing correct subsystems. ### Source excerpt Back when I was an undergraduate, as part of a class called "Computer Systems Engineering", we read numerous classic papers of systems design. I enjoyed and learned a great deal from many of these papers, but one that paper that has stuck with me in particular was Saltzer et al's "End-to-End Arguments in Systems Design". The paper is a very general tract on systems design - it does explore several examples of concrete systems or applications, but it ultimately expounds upon the end-to-end principle as a perspective or design heuristic that can apply to virtually any system design. ## Speaking at Tech Mesh DevFeed: [Speaking at Tech Mesh]() Original publisher: [Read original article]() Author: Michael Barker (noreply@blogger.com) Published: 2012-11-07T23:15:00Z Content type: opinion Language: en Sources: [Bad Concurrency]() Topics: [Development](), [systems]() Tags: [announce](), [challenges](), [concurrent](), [disruptor](), [hardware](), [high-performance](), [history](), [jvm](), [performance](), [perspectives](), [speaking](), [systems](), [tech](), [time]() ### AI overview The author announces an upcoming Tech Mesh talk about the Disruptor, covering its history and motivations, the challenges of building high-performance concurrent systems, and how JVM and hardware changes could support them. ### Source excerpt I'm happy to announce that I will speaking at Tech Mesh in December. I'll be speaking about the Disruptor from two perspectives, firstly looking briefly back at some of the history and motivations behind the Disruptor. Then spending some time explaining at the challenges of building high performance concurrent systems (like the Disruptor) and delving into how the JVM and hardware could change to support the development of these systems.