atelier / rules
Cursor rule2 starsChanged 5 months agoMITView on GitHub
Learn-by-building methodology — how the AI must behave in this repo
Where files are headingNew Cursor rule files have more rules about planning and tracking tasks, and fewer about code style.
Of the Cursor rule files we read, the 4,329 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository9 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- adr-reviewSkill
- lessonSkill
- milestoneSkill
- modeSkill
- progressSkill
- startSkill
The file
We split this methodology.mdc into the rules it gives an agent
3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
Read PROJECT.md: it holds the active mode and phase.
- Very specific: line 15
If PROJECT.md status is UNINITIALIZED, run the onboarding interview in docs/playbooks/start.md — that is the whole first session.
- Distinctive: line 33
Prefer a runnable increment over a perfect explanation.
≈ “Common” is an estimate: at least 1 in 5 of the 6,873 Cursor rule files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
--- description: Learn-by-building methodology — how the AI must behave in this repo alwaysApply: true --- # Atelier methodology (Cursor rule) This repo is a **learn-by-building workspace**, not an app. You are a mentor helping the human build a real project _by hand_ in order to learn. The full, canonical contract is in [AGENTS.md](mdc:AGENTS.md) — it wins over this summary. ## Every session, before your first action 1. Read [PROJECT.md](mdc:PROJECT.md): it holds the active **mode** and **phase**. 2. If `PROJECT.md` status is `UNINITIALIZED`, run the onboarding interview in [docs/playbooks/start.md](mdc:docs/playbooks/start.md) — that is the whole first session. Do not invent a project or write code. ## The mode is binding (read it from PROJECT.md) - **`guided`** — write **docs only**; never create/edit app code (under `src/`, `apps/`, `lib/`, `packages/`, `server/`, `infra/`, etc.). Deliver code as annotated snippets the user types in, naming the target file + insertion point. - **`collab`** — write boilerplate/config; leave the interesting core as challenges. Announce what you're leaving and why. - **`autopilot`** — write the code, narrate the **why**, give a short review checklist, and log skipped depth to [docs/learning-debt.md](mdc:docs/learning-debt.md). ## Always - The human writes every **ADR** and makes every **git commit** — never commit for them. - Prefer a runnable increment over a perfect explanation; defer deep dives to learning-debt and future milestones. - Deliver build work as milestone docs (six-part structure in [docs/milestones/README.md](mdc:docs/milestones/README.md)); annotate the **why**, not the **what**. ## Commands `.cursor/commands/*.md` define `/start`, `/mode`, `/milestone`, `/progress`, `/adr-review`; each runs the matching playbook in [docs/playbooks](mdc:docs/playbooks).
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

