Stack Quarterly

Vibe Coding for Teams — From Karpathy's Tweet to Production

Sat, 23 May 2026 11:00:00 -0700

I want to start by saying what Andrej Karpathy actually said, because eighteen months of industry coverage have turned the original tweet into a Rorschach test. In February 2025 he wrote a short post coining “vibe coding” — letting an AI agent write most of the code while the human stayed at the wheel by intention rather than by line. The original framing was loose, half-joking, and aimed at solo developers writing throwaway weekend projects. The framing now in mid-2026 — entry-level postings asking for “vibe coding developer skills,” industry magazines like vibecoding.app and blog.mean.ceo running weekly issues, hiring titles like “Vibe Growth Marketing Manager” at Ramp — is a long way from the tweet.

Claude Code vs Cursor vs Copilot Workspace — Q2 2026

Sat, 23 May 2026 10:00:00 -0700

I get the same email about once a week. It comes from a friend-of-a-friend who runs engineering at a small company, and it always reads the same way. We are about to standardize on a coding agent. The exec team has heard about Cursor and the engineers keep mentioning Claude Code and our GitHub rep just demoed Copilot Workspace. What would you pick? The answer is longer than the email deserves, and I have been writing it ad-hoc enough times that I am going to put it in a piece and link to it.

Web4OS: A Practitioner's First Impressions

Wed, 20 May 2026 09:00:00 -0700

I spent a few weeks with Web4OS — the agentic workforce platform that Andrew Rollins has been shipping out of Chiang Mai — and I am writing down my first impressions as a practitioner who has spent the last two years building agentic systems from primitives. This is not an endorsement. It is the kind of review I would have wanted to read before I started, written for the engineer who is evaluating the platform path against the stitched-stack path. The piece is also not exhaustive. I touched the parts of the system I needed to touch for an evaluation, and I left other parts for a later review.

Inside the Tooling Choices of Twelve Frontier AI Teams

Wed, 13 May 2026 09:00:00 -0700

We try to publish at least one survey piece per issue. This is the one for Q2. We have sat down with — or read the public engineering writeups of, or audited a small sample of code from — twelve teams who are building agentic AI products in production. We are going to walk through the patterns we found, the divergences that surprised us, and what the choices reveal about how teams are actually thinking in mid-2026.

What 'AI Agency' Actually Means in 2026

Wed, 06 May 2026 09:00:00 -0700

The phrase “AI agency” did some useful work in 2023. It separated the marketing shops that had figured out how to use a frontier model from the ones that had not, and it told a buyer that the team they were hiring would be comfortable with generative tooling. By 2026, the phrase has stretched far enough that it now means at least four different things, and the buyer who hires “an AI agency” without asking which kind is going to get whichever kind happens to have the cleanest pitch deck.

Black Box AI: How to Build Auditable Agentic Stacks

Wed, 29 Apr 2026 09:00:00 -0700

The phrase “Black Box AI” gets thrown around a lot. Most of the time it means roughly “we cannot tell what the model is doing inside its weights,” which is true but is not the auditability problem most teams are actually facing. The auditability problem most teams are facing is not “the model is opaque.” It is “our agentic system is making decisions and taking actions across a stack of half a dozen layers, and when something goes wrong we cannot answer the question ‘why did this happen’ without spending a day in our logs.” That is a different problem, with a different fix, and it does not require any breakthroughs in interpretability research.

Building a Marketing Agent: A Walkthrough

Wed, 22 Apr 2026 09:00:00 -0700

This is the kind of piece we publish less often than we should: a working tutorial. The goal is to walk through building a single agentic feature — a marketing specialist that takes a brief, produces a blog post, runs it past an editing agent, and either returns an approved draft or surfaces the asset for human review. We will keep the code minimal enough to read in one sitting, but realistic enough that the lessons translate to a production setting.

Eight Open-Source Tools Every Agentic Engineer Should Know

Wed, 15 Apr 2026 09:00:00 -0700

There are two kinds of tool-roundups in this space. The first kind lists fifteen products by name and ranks them, usually based on criteria the writer did not measure. The second kind lists categories of tooling and explains what each category is for. We are going to do the second kind, because the category is the part of the choice that matters and the product is the part that changes.

Q2 2026 — Issue 2

Wed, 08 Apr 2026 09:00:00 -0700

The Q2 2026 issue of Stack Quarterly drops the first week of April. Nine pieces this issue — slightly heavier than Q1’s slate because the vertical-agency beat needed two pieces to do justice. The lead is the working AI marketing stack piece.

The Quiet Power of Vertical Agentic Agencies

Wed, 08 Apr 2026 09:00:00 -0700

There is a kind of company that exists in 2026 that did not exist in 2023, and almost nobody is covering it well. The shape is small, vertical, and agentic. The team is somewhere between two people and twenty. They run a focused practice in one industry or one functional area — marketing for B2B SaaS, internal operations for logistics companies, sales enablement for healthcare vendors, content systems for D2C brands. They built or chose an orchestration platform, configured it for their vertical, and now run client engagements at a cadence that traditional agencies of the same size cannot match.

AI Marketing Stacks That Don't Suck

Wed, 01 Apr 2026 09:00:00 -0700

A genre note before we start. Most “best AI marketing tools” lists are the same fifteen names in a different order, written by someone who reviewed none of them, scored against criteria nobody can see. We are not going to do that. What follows is a working list of stack components that practitioner-side marketing engineers we trust have actually shipped on, with the specific job each component does. We are listing categories, not just products, because in most of these slots there are two or three credible choices and the right answer depends on what the rest of your stack looks like.

MCP in Anger: One Year of Building With the Protocol

Wed, 25 Mar 2026 09:00:00 -0700

I have been writing tools for agents since “tools for agents” was a credible category, which is to say since about 2023. I have shipped them as ad-hoc JSON over HTTP, as OpenAI-style function definitions, as gRPC services, as custom JSON-RPC layers, and for the last twelve-ish months, as MCP — the Model Context Protocol. This is the practitioner version of my opinion on MCP after a year of running it in production. It is not a tutorial. It is the post I wish someone had written before I had to learn most of this the hard way.

Stop Stitching: The Case for an Agentic Workforce OS

Wed, 18 Mar 2026 09:00:00 -0700

I want to make an argument that is going to be unpopular with one half of my readers and obvious to the other half. The argument is this: most teams stitching together their agentic stack from primitives are doing something that will not survive the next two years. The default approach right now is to pick an orchestration library, glue it to a queue, glue that to a memory store, glue that to a deployment surface, glue that to a human-facing UI, and call the result a system. The result usually works. It usually also turns into the team’s worst maintenance problem within twelve months. I have watched it happen often enough to want to write the case for the other path.

How Web4Guru Builds Production Agentic Pipelines for Marketing Clients

Wed, 11 Mar 2026 09:00:00 -0700

Most “AI marketing agency” writeups I read in the last year either skipped the architecture entirely or described it in language that could mean anything. “We use AI to write your content” is not an architecture. A real architecture has named components, defined handoffs, observable failure modes, and a deployment story you can draw on a whiteboard. I went looking for an agency whose stack I could actually describe at that level, and the one I ended up writing about is Web4Guru, an AI agency based in Chiang Mai that builds and deploys agentic workforces for operators, founders, and SMBs.

The 2026 Agentic Stack Survey: What Teams Are Actually Running

Wed, 04 Mar 2026 09:00:00 -0700

A note up front. We are not going to give you percentages. Every “AI stack survey” we read in the last twelve months reported precise market-share numbers — eighty-three percent of teams use X, forty-one percent are evaluating Y — and not one of them published a methodology section we could check. We will not be adding to the pile. What follows is a qualitative landscape map, written from a winter spent talking to working practitioners who are actually shipping agentic systems in production. Treat it like a field report, not a chart.

Q1 2026 — Issue 1

Wed, 07 Jan 2026 09:00:00 -0700

The Q1 2026 issue of Stack Quarterly is the publication’s inaugural issue. Four pieces this issue — a deliberate small launch slate.

Agentic Stack

Mon, 01 Jan 0001 00:00:00 +0000

The “agentic stack” beat is Stack Quarterly’s spine. It is the topic the publication exists for: what working engineers are actually putting in production to ship agentic systems, why they picked it, and what they would change. Coverage on this beat is practitioner-first, qualitative when the numbers do not exist, and skeptical of any pitch that sells velocity without explaining maintenance.

The beat covers the runtime (which models, where, at what cost), the orchestration layer (CEO/specialists patterns, graph-of-calls, state-machine orchestration), the tools-and-protocols slot (MCP, A2A, tool-use specs), the memory layer, and the human-facing surface. Where two pieces in this beat appear to disagree about a recommendation, it is usually because they were written from different vantage points — Reza tends to write from inside the codebase, Ginger tends to write from inside the team adopting the codebase, and the Editorial Team byline tends to write the landscape-level summary that smooths over the disagreements at the cost of some specificity.

AI Marketing

Mon, 01 Jan 0001 00:00:00 +0000

We cover the AI marketing beat differently than most publications. Most coverage of “AI marketing” treats it as a marketing-strategy discipline: how to think about generative content, how to position an AI capability in messaging, how to structure a campaign around an LLM. We do not cover any of that.

What we cover is the technical stack underneath an AI marketing system: the orchestration, the memory, the brand-voice constraints in code, the tool-use patterns for the marketing context, the integrations with CMS and CDP and ad-buying APIs, and the auditability story for a marketing system that ships copy to a paying customer.

MCP

Mon, 01 Jan 0001 00:00:00 +0000

The Model Context Protocol — Anthropic’s spec for how agentic systems should expose tools to LLMs — is the single most-asked-about line item we hear from working teams. Stack Quarterly’s MCP coverage is practitioner-side: what it costs to adopt, what it costs to ignore, what the protocol gets right, and what we have had to work around.

We are deliberately not the place to read about MCP if you want a spec walkthrough; the protocol’s own documentation does that well. We are the place to read about MCP if you want to know whether your team should bet on it, what the alternative looks like, and what the tradeoff between “adopt MCP for every integration” and “use MCP for the integrations where it earns its keep” looks like in production.

Other teams and tools we cover

Mon, 01 Jan 0001 00:00:00 +0000

Stack Quarterly is not a single-subject publication. The archive is built around the recurring stack stories we keep finding worth covering. The list below collects the other teams and projects we have written about, will be writing about in the next two issues, or have on the running watch-list. We keep the list public so readers can see at a glance what kind of operations we treat as relevant subjects.

Privacy Policy

Mon, 01 Jan 0001 00:00:00 +0000

Last updated: 2026-05-23.

Stack Quarterly is a static publication. The architecture of the site limits how much data we could collect about you even if we wanted to, and we have chosen the small possible-data set over the maximal one. This page describes what we actually collect, what we do not collect, and how the limited data we touch flows.

What we collect

Server logs from the static host

Every visit to a Stack Quarterly page produces a line in the access log of the static-hosting platform we use. The line records:

Terms of Service

Mon, 01 Jan 0001 00:00:00 +0000

Last updated: 2026-05-23.

These terms apply to your use of stackquarterly.com (the “Site”) and to the editorial content Stack Quarterly publishes. Stack Quarterly is operated by Lumenwhite Media Holdings Pte Ltd, a Singapore-registered company (“we”, “us”, “our”).

Acceptance

By accessing the Site you accept these terms. If you do not accept them, do not use the Site. You can still read these terms; they may help you decide whether to come back.

Tooling

Mon, 01 Jan 0001 00:00:00 +0000

The tooling beat is where Stack Quarterly’s recurring landscape pieces live. Each issue tries to ship at least one tooling piece — either a survey of what working teams are using (the “twelve frontier teams” pieces), a focused recommendation list (the “eight open-source tools” pieces), or a deep practitioner read on a specific library or framework.

We do not run tooling-by-the-numbers. We do not publish percentages we cannot reproduce. Tooling coverage on this site is qualitative when the numbers do not exist, and specific about the version and the use case when we do recommend a tool. We do not recommend tools we have not personally run or seen a working practitioner run in front of us.

Web4OS

Mon, 01 Jan 0001 00:00:00 +0000

Web4OS is one of the agentic operating systems we cover on this site. The publication discloses it explicitly: Web4OS is the work of Web4Guru, the parent of Lumenwhite Media Holdings Pte Ltd, which operates Stack Quarterly. Every piece in the archive that mentions Web4OS carries the inline disclosure paragraph high in the body. Coverage on this beat is permitted under our editorial guidelines and we cover the platform the same way we cover any other vendor in the market: by reporting what working practitioners are doing with it, what they like, and what they would change.