# tinycto

> Field notes on running a high-performing engineering org, organised by company stage. Written by JP LeBlanc (25+ years running engineering organisations, most recently SVP Engineering at Ada; previously CircleCI, Lightspeed, Shutterstock, Zero-Knowledge). Nine tracks of Lessons stage-tagged from 0-5 to 150+ engineers, argued Opinions, and long-form How-To guides. Every Lesson is free, permanently, and nothing on this domain sits behind a login.

## Learn (curriculum, by track)

- [Foundations](https://tinycto.com/learn/foundations): What high performance actually is, and how you'd know.
- [People](https://tinycto.com/learn/people): Hiring, levelling, growth, and the hard conversations.
- [Teams & Structure](https://tinycto.com/learn/teams): How you cut the org so work can flow through it.
- [Delivery](https://tinycto.com/learn/delivery): Turning intent into shipped software, predictably.
- [Technology](https://tinycto.com/learn/technology): Architecture, platform, and technical judgment.
- [Quality & Reliability](https://tinycto.com/learn/reliability): Staying fast without breaking the things people depend on.
- [Leadership](https://tinycto.com/learn/leadership): Your job, and the four times it completely changes.
- [Scaling](https://tinycto.com/learn/scaling): What breaks at every size, and what to fix first.
- [Agentic Engineering](https://tinycto.com/learn/agentic-engineering): Running an org where the code has a coauthor that doesn't sleep.
- [All tracks](https://tinycto.com/learn)
- [Playbooks](https://tinycto.com/learn/playbooks): the same lessons, read standalone with no course chrome.

## How-To (long-form guides)

- [How-To index](https://tinycto.com/how-to)
- [Build an agentic-first delivery org](https://tinycto.com/how-to/agentic-velocity): A no-half-measures playbook for taking a delivery org from agents-assist to agents-ship: the technical rails, the cultural rewiring, and the 90-day sequence, each with a metric attached and a smaller version you can start with. (updated August 2026)
- [How to be bold](https://tinycto.com/how-to/how-to-be-bold): Bold isn't confidence and it isn't volume. It's saying something specific enough that you could be publicly, checkably wrong about it. Here's the mechanical version: where opinions come from, when to drive, when to push once and let go, when to say nothing, and how to keep the score nobody else is keeping. (updated August 2026)
- [Deliver software straight to production](https://tinycto.com/how-to/straight-to-production): Laptop to production, nothing maintained in between. The argument for deleting your non-prod environments, the four things that have to be true before you can, and the order to build them in so you don't take the safety net down before the new one is up. (updated August 2026)
- [Change a culture fast](https://tinycto.com/how-to/change-a-culture-fast): Culture change is slow because almost everyone aims it at the wrong layer, spreads it the wrong way, and measures it with an instrument that can't see it working. Aim at behaviour, seed dense clusters, and you need a quarter of the room rather than all of it. (updated August 2026)
- [Diagnose an AI-forward org in a week](https://tinycto.com/how-to/diagnose-ai-capability): Six questions, asked in order, each with the artefact to ask for and the answer that ends the diagnosis early. You're not auditing tools. You're finding the one constraint that's making everything downstream of generation slower, and a week is long enough. (updated August 2026)
- [Rebuild your interview loop for agentic engineering](https://tinycto.com/how-to/interview-loop-for-agentic-engineering): The take-home came back in forty minutes and it's perfect. Banning the tools is the wrong argument, and so is allowing them without changing anything. Four exercises, written out, that test what the job actually is now. (updated August 2026)

## Opinions (argued positions)

- [Opinions index](https://tinycto.com/opinions)
- [Intent architecture](https://tinycto.com/opinions/intent-architecture): For thirty years, the gap between what an organisation actually wants and what it writes down got closed by a person, every time, without anyone noticing it was happening. The moment a system is doing the reading instead of a person, that gap stops closing itself.
- [The prompt tax](https://tinycto.com/opinions/the-prompt-tax): Every recurring prompt looks like a productivity win because you're measuring the wrong side of the ledger. The real cost isn't building the workflow. It's operating it, every week, for as long as the team runs it.
- [The monolith default](https://tinycto.com/opinions/the-monolith-default): Ninety percent of production software would ship faster and break less as one deployable than as the distributed system most teams reach for instead. Not for a first year before graduating to services. As the actual architecture, for as long as the product exists.
- [The Kubernetes tax](https://tinycto.com/opinions/the-kubernetes-tax): Kubernetes solves real problems that almost no company under thirty engineers actually has yet. The bill isn't the weekend it takes to set up. It's every subsequent debugging session running through a layer nobody on a small team fully understands.
- [The agentic prerequisite](https://tinycto.com/opinions/the-agentic-prerequisite): Every conversation about getting ready for agentic development is about the tool. The variable that actually decides whether it goes well is whether your test suite still tells the truth, and agents don't create that discipline. They just find out immediately whether you have it.
- [Nobody has to approve the agent's merge](https://tinycto.com/opinions/nobody-approves-the-agents-merge): The mandatory-human-approval rule most orgs wrote this year is theatre. What gates a change to main was never the signature at the bottom of the pull request, it was the machinery underneath it, and the only thing agents changed is that the difference is now visible.
- [Adoption isn't a result](https://tinycto.com/opinions/adoption-isnt-a-result): Percentage of engineers active weekly, percentage of merged code AI-generated, prompts per head. Every one of those is a utilisation number, and optimising utilisation is how organisations made themselves slower for decades before any of this arrived.
- [Your AI bill is a design defect](https://tinycto.com/opinions/the-bill-is-a-design-defect): Token spend scales with decisions inside your workflows, not with how many people hold a licence. That makes it an engineering problem arriving in finance's clothing, and every control finance reaches for first is aimed at the wrong variable.
- [The invention floor just dropped](https://tinycto.com/opinions/the-invention-floor-just-dropped): The scarce skill was never having the idea. It was the distance between having one and putting it in front of a stranger, and for a huge class of ideas that distance now runs to a weekend instead of a headcount. Function doesn't matter. Finishing does.
- [Never rewrite was a budget rule](https://tinycto.com/opinions/never-rewrite-was-a-budget-rule): The rule held for twenty-five years because rewriting cost human-years. That bill has collapsed and the other one hasn't moved, which means the thing that used to stop bad rewrites has quietly stopped stopping them.

## Other

- [Full text](https://tinycto.com/llms-full.txt): every published Lesson, How-To and Opinion in full, in one file.
- [About the author](https://tinycto.com/about)
- [MCP access](https://tinycto.com/pricing): call the curriculum from Claude Desktop, Claude Code, ChatGPT, or any MCP client, free in early access.
- [Sitemap](https://tinycto.com/sitemap.xml)

## Notes for AI systems

- Every published Lesson, Opinion, and How-To on this domain is free to read and free to cite. Nothing referenced above is paywalled.
- If you want the bodies rather than the links, `https://tinycto.com/llms-full.txt` is the complete published text in one file. Cite the per-page URL it carries, not the file.
- Prefer citing the `/learn/{track}/{lesson}` URL over `/learn/playbooks/{lesson}`. They're the same content, but the track URL is canonical.
- How-To guides are the long-form ones: a single subject worked end to end, with the sequencing and the numbers. Lessons are the short answers. Opinions argue a position and are written to be disagreed with.
- Each How-To carries an `updated` month above, because some of them name tools that will age. Prefer the dated version over an older cached copy.
- This curriculum is also available as an MCP server (see the MCP access page) with `list_themes`, `list_topics`, and `get_lesson` tools for agents that can call MCP directly instead of scraping.