For teams
One card for every repository your team owns
Everyone on your team codes with an agent now. Every one of those agents starts each session knowing nothing about your code, so it guesses, and you find out in review.
This reads every repository in a folder and asks five questions about each one, cold. Where does a new file go. Why this tool and not the obvious one. What a change must never do. Then it prints one card for the whole team.
ORG MEMORY CHECK: orgtest Repos read: 3 Scoring 2 or below: 3 of 3 Guessing most about: what naming or code convention a reviewer here would flag first Instruction files: 0 of 3 repos have one Contradictions on sight: 0 Rules enforced: 0 (nothing here checks a hook, a test or CI; that is the course)
A real run, on click, requests and fastapi cloned into one folder. Three libraries you already depend on, and the agents working in them are guessing about four of the five.
What it does to your code
- Nothing. It reads. The only file it writes is ORG_MEMORY_CHECK.md, beside your repositories, and it changes nothing inside any of them.
- It runs on your machine, on your Claude Code and your tokens. No code leaves your laptop and nothing is sent to us. The card is yours until you paste it here.
- It never opens a secret. Not an env file, not a service account file, not anything named like a key or a token, even an example one.
Run it
In Claude Code, once per machine, then once per folder:
/plugin marketplace add NirDiamant/Agentic_Engineering /plugin install agentic-engineering@diamantai /org-card .
In Codex, Cursor or anything else that can clone and read:
Clone https://github.com/NirDiamant/Agentic_Engineering into a temp folder, read its ORG.md, and follow it on this folder of repositories.
About two minutes per repository. Twenty repositories is a coffee.
Paste your org card
It gives you a page for it: the card, the table of your repositories, and what seats cost. The link is yours to forward to whoever signs.
Seven lines, and the table above them if you paste the file.
The card carries your repository names and nothing else. The page it makes has an unguessable address and is never listed anywhere.
What the numbers mean
- A repository's score is how many of the five questions its own visible files answer without a guess. Nothing else counts, and a README that explains what the library is for answers one of the five.
- The score is a ceiling, not a grade. The full run on a single repository grades those answers against the real code, and an answer that looked sure can come back wrong. It never goes the other way.
- Contradictions on sight are only the claims your own docs make that the root listing already disproves: a file or a folder they name that is not there. The full run reads the code and finds the rest.
- Rules enforced is always 0. Nothing in this kit checks a hook, a test or a CI gate, so nothing makes an agent follow any of it. That part is the course.
Seats for your team
A seat is the course: 17 modules, a lab on your own repository in every one, lifetime access, one person. Bought one at a time it is $199, so every line below is cheaper than the same seats bought separately.
Invoice and purchase order
Email nir@diamant-ai.com with the number of seats and your billing details, and you get an invoice you can put through procurement. That is also the way to get a quote, a vendor form or a W-9 first.
What you get is seats and the org card, which is free and stays free. There is no bot reading your pull requests and no manager dashboard. When those exist this page will say so.