From the editor

A bet on the next decade.

Three claims about what's broken, what changed, and the shape of the next decade of company-building software. The argument behind YAGNI.

The thesis By Jack Collins · Founder Denver · 2026

The argument, in three claims

Three things have to be true.

Information overload at fifty.

When you have five tools, each knows its slice and you keep the rest in your head. When you have fifty, the rest doesn't fit anywhere, not in any one tool, not in any one person, not in any one Monday standup. The cost flips from per-seat licences to a tax on attention.

By the time a 75-person company is running honestly, it has Notion AI for the doc layer, Granola for the meetings, Gong for the calls, Superhuman for the inbox, half a dozen point-AI features bolted onto Slack, Linear, HubSpot, and a dashboard for every function that nobody opens. Each one knows a sliver. None of them know the whole thing. The status meeting comes back anyway.

The fix isn't another tool. The fix is to stop trying to compose status in human meetings, in slide decks, in the hour before all-hands, in the weekend before the board call. The fix is an editor.

II.

AI is a teammate, not a feature.

The instinct in 2024 was to add an AI sidebar to every existing tool. Notion AI. Gmail AI. Slack AI. The sidebar was a useful first cut. But the sidebar is the wrong shape: it asks you to sit inside the tool you opened, hoping the AI knows enough about that one tool's slice to be useful.

The right shape is a teammate. One identity, one memory, every doorway. Slack mention, Gmail label, dashboard, repository, calendar invite: the same agent, knowing your business across all of it. Models in 2026 are good enough to be that teammate. Sidebars in your existing tools were never going to be.

We took the bet on the teammate shape early. The product is a coherent agent surface, not an AI add-on. One @yagni per workspace, reachable from anywhere your work already lives, with the same memory regardless of doorway.

III.

An Edition beats a dashboard.

A dashboard waits for you to look. An Edition decides what's on the front page. It has voice. It has an editorial line. It curates. It writes for you, not at you. The Tuesday paper opens with the right lead, not because you queried for it, but because the agent read the week and made the call.

We chose the newspaper shape because it's the right shape for an agent that reads everything and tells you what matters. Editors decide what to publish, what to spike, what to chase down. Editors care about voice. Editors know the reader. None of those words apply to a dashboard. All of them apply to what a 75-person company actually needs.

Underneath the metaphor is real machinery: @yagni reads your tools continuously, Field Work runs investigations on commission, and approved drafts turn into action. But the metaphor isn't a stunt; it's the right organising shape. You read the paper. You hit approve. The work goes out.

"The destination is agent-completed business work. The wedge is the Edition."

What we sell, what we don't

One thing, not many.

We sell

The agent.

The Edition, and the agent that ships work on it. That's it. That's the whole product. Across coding, research, reports, outbound, customer setup. One agent. One memory. One paper.

We do not sell

Anything else.

Not hosting for your app. Not a dashboard for any single tool. Not a team chat replacement, a CRM, a project tracker, a knowledge base, an analytics product, a no-code automation tool, or a general-purpose AI assistant. We read every tool you already pay for. We don't replace any of them.

A note from the editor

Why we're building this.

By Jack Collins · Founder · Denver · 2026

I started YAGNI because by the time I was running a fifty-person company I was paying for seventeen different software subscriptions and operating none of them well. Each tool knew its slice. The slice it knew was good. The thing nobody knew (what I should do today, what mattered most across all of it, what was on fire and what was quiet) lived in my head, leaking, and on a Slack standup that took an hour and produced a list of action items I half-finished.

The instinct everyone in 2024 had was to add AI to one of those seventeen tools. Notion AI for docs. Gong AI for calls. Slack AI for threads. None of those are wrong. But none of them are composable: none of them write a single Tuesday paper that says, plainly, here's what matters today, here's what to do about it, here's the draft for the part you'd write yourself.

I think the next decade of company-building software belongs to a single agent that reads every system you use, composes today's Edition, and ships the work you approve. You read it. You hit approve. The work goes out. That is what an editor would do for a publication, and a 75-person company is a publication: it has audiences, it has beats, it has a voice, it has a Tuesday morning.

We're a small team. We're shipping fast. We sell one thing. We'd like you to subscribe.

Jack

Founder · jack@yagni.app · Book a call

Run it across your team.

A demo covers your stack, the Standard Desks that fit your org, the rollout sequence we've watched work. Forty-five minutes with the founder.

Book a call → Start free or read today's edition