--- title: "Work: AI systems that give you back the week" description: "Four AI systems built by Control Square: a publishing harness, an agent-readable publication, a book and curriculum canon, and a token-frugal architecture." url: "https://controlsquare.com/work/" --- # AI systems that give you back the week. Control Square builds AI systems that take the repeatable work off a small business and leave the judgment where it belongs. Four builds, one pattern: use AI where it earns its place, keep the human decision, leave proof. ## Key facts - 3 issues live on a homeschool publication - 48 posts readable by AI agents - 26 books planned, 17 drafted (A to Q) - 4 token rules, 0-token scripts, 28 converter tests ## 1. An AI publishing harness for a solo education business **Status:** AI system, live. A solo publisher can hold the thesis and still lose the week to everything around it: finding a topic with demand, checking the angle, sourcing claims, generating images, formatting for Substack, and keeping the voice consistent. We built an AI harness that does that work and stops at the decision. It researches demand and seasonality, drafts in the publisher's voice, checks structure, sourcing, and layout, generates and optimises images, and leaves a formatted draft in Substack. AI handles the open-ended work; scripts handle the repeatable work and the guardrails. Missing positioning stays missing, invented experience fails, and a 22-check guard blocks writes to the wrong publication on a shared account. Publishing stays a human action. The live publication is at [forgotteneducation.substack.com](https://forgotteneducation.substack.com). **Case study:** [Inside a token-frugal publishing harness](https://controlsquare.com/work/ai-publishing-harness) **Proof:** 3 issues live, 22 guard checks, one decision per post. ## 2. A publication that AI agents can actually read **Status:** AI infrastructure, 2026. Most sites answer an AI agent with a full HTML page, a cookie banner, and a wall of navigation. We built markdown negotiation into this publication: when a request asks for `text/markdown`, the site returns the article with its frontmatter, and advertises the alternate in the page head and HTTP headers. One site now reads cleanly to people and machines: 48 posts converted, Level 3 Agent-Readable in an external scan, with tables, code, captions, and nested lists preserved. HTML stays the default for everyone else, and one plugin covers every post instead of a hand-built sample. **Proof:** Level 3, 48 posts, one plugin. ## 3. One canon, three products: books, curriculum, shop **Status:** Build, in progress. A 26-book alphabet series, a homeschool curriculum, and an Etsy shop can drift into three businesses. We built a file-based operating system that keeps them one: locked canon for the series, one brief per project, and a shared voice that survives every new letter, lesson, and listing. The system generates art with image models, styles it, builds print-ready PDFs, and turns finished units into shop products. The publication is the front door we are building; the books and curriculum are the deeper offer. Same owner, same business, one source of truth. **Proof:** 26 books planned, 17 drafted (A to Q), one canon. ## 4. An AI harness built to spend tokens like money **Status:** Architecture, shared across the builds. Most AI systems get expensive because every step pastes its work back into the conversation. We built the opposite. Subagents write to files and return a path, state lives in files, and only judgment calls reach a model. Formatting, resizing, scheduling, and tests run as scripts. If a script can do it, a model never sees it. The same architecture guards quality. Sources are checked, voice drift is measured, layout is gated, and a human makes the publish decision. It costs less to run and is harder to degrade as it grows. Cheap is not the point. Durable is. **Proof:** 4 token rules, 0-token scripts, 28 converter tests. **Reference implementation:** [github.com/ControlSquare/token-frugal-harness](https://github.com/ControlSquare/token-frugal-harness) ## The pattern ### Automate the mechanical. Research pulls, image resizing, formatting, and guard tests run as scripts. AI is used for the open-ended work, not for what code does better. ### Keep the judgment human. Topic, argument, voice, publish. One decision per piece, not a dashboard to babysit. ### Proof over promises. Tests, scans, and live artifacts. If it cannot be checked, it does not go on this page. ## Start here Tell us where the work piles up. We will tell you what AI can take, what it should not, and what it would take to build. Contact: https://controlsquare.com/contact/ Some builds are internal or partner work. Names are withheld where the work is private. The systems are real.