Quick start

One line to your agent. There is nothing to install and nothing to build.

follow repositorystandards.com - take this repo onto the standard, interview me for what you need

That is the whole thing. The agent reads the standard, works out whether your repo is new or has ten years of history, and asks you what it cannot work out for itself.

The other three lines you will ever need

You want the number before committing to the work. Nothing is changed:

score this repo against repositorystandards.com - count the work, do not do it

Later, when the standard has moved. It is a delta, not a re-scaffold, and your recorded deviations survive it:

update me to the latest repositorystandards.com

You are on a registered stack. Both layers, one number:

follow repositorystandards.com - adopt the standard with the Node stack

What you get back

A tree in your repository: specs by capability, decision records, a persona roster, a backlog that feeds itself, and scripts/ - so from then on the claim is checkable by anyone, including your CI:

node scripts/self-verify.mjs

It exits non-zero on any failure and prints drift as a number. That number is the contract: not "we follow a standard", but "we are this far from it, and here is the list".

If your agent cannot browse

Fetch the standard once and point at the checkout instead. It is a cache you read, not a dependency you vendor, so add it to .gitignore:

npx degit repository-standards/core .repository-standards

follow .repository-standards/skills/align-to-standards/SKILL.md - take this repo onto the standard

You need Node and jq on the machine for the shipped guards to work. Without jq the pre-execution guards deny every command rather than pass it unchecked.

If your agent is not Claude Code

The skills ship in Claude Code's format. Port them to your agent's own mechanism - strictly, not approximately, because a skill that paraphrases the loop is a loop that quietly does something else.

Now find yourself

The loop looks different depending on which hand you are. Read the one that is yours, not all three.

	go here
You decide what the product should do	Product Owner - the four answers you are allowed to give, and what you must never be asked to do
You build it	Developer - raising a spec to buildable, the guard, and reading a refused plan as your to-do list
You are rolling this out across a team or a client	Consultant - the order to do it in, the three objections, and the two ways it goes wrong
Any of the above, mid-task	Anyone - find your situation, say the line

Want the whole picture first? Start here walks the loop once, end to end. Wondering what the agent handles without being told? What it does by itself.