MADEBYARIS
From the first Cursor Ambassador in Indonesia

Cursor guide for teams that ship

Cursor works on a real repo when three things are in place: half a page of project rules, a plan in chat before the agent writes anything, and a human review of every diff. This page puts my posts on each step in order, with the rules I commit and a cost estimate for your team.

The loop I run on client work

  1. Step 1

    Rules

    Half a page that names the stack, the folders, and what not to touch.

  2. Step 2

    Plan

    Ask chat where the change lives. Real paths, no files written yet.

  3. Step 3

    Scoped run

    Point the agent at those files only, with a blast radius in the prompt.

  4. Step 4

    Review

    Read the diff like you own the on-call. Then test and commit.

Reading order

18 posts, grouped by where you are. See them newest first.

Start here

Set up Cursor on a real repo and learn the loop I use on client work.

Rules, context, and setup

Project rules, MCP servers, and picking the right surface for the task.

Shipping and review

Keep AI-written code reviewable so it still looks like your codebase next month.

Teams, cost, and tool choice

Rolling Cursor out to a team, what it costs, and how it compares.

Bahasa Indonesia

Panduan Cursor dalam Bahasa Indonesia.

Rules you can copy

Drop these into .cursor/rules/ and edit the paths to match your repo. Each one comes from a post that explains why every line is there.

Project rules for a Next.js app

Stack, folder map, error shape, naming, and test command on half a page. Stops Cursor inventing packages and folders.

---
description: Project rules for a Next.js App Router repo
alwaysApply: true
---

# Project rules

Stack: Next.js App Router, TypeScript, pnpm, Tailwind.

Folders:
- Routes: app/
- UI primitives: components/ui. Reuse these. Do not invent a new button.
- Shared helpers: lib/
- Do not create a new package or a new top-level folder.

If a helper is missing, ask before creating it. Prefer lib/ over a new utils/.

API errors: { error: string, code?: string }. Match the nearby route handler.

Naming: kebab-case files, PascalCase components. Match the sibling file.

Tests: pnpm test. Colocate as *.test.ts.

Do not change schema, auth, or payments unless the prompt names those files.
No drive-by refactors.
DownloadWhy each line is there

Review checklist for AI-written code

The checklist I paste into every PR that an agent drafted. The agent runs it before it calls the work done.

---
description: Run this checklist before you call a change done, and paste it into the PR description
alwaysApply: false
---

### AI-Assisted Code Verification Checklist

- [ ] **Boundary Check:** All modified files are strictly within the intended feature scope.
- [ ] **No Unvetted Deps:** No unapproved packages added to package.json / lockfile.
- [ ] **Deduplication:** Verified that existing helpers and types were reused rather than recreated.
- [ ] **Security & Auth:** Server-side input validation and auth checks are explicitly tested.
- [ ] **Error Handling:** Empty, loading, and failure states are tested and accounted for.
- [ ] **Real Tests:** Unit/integration tests verify business outcomes, not just mock tautologies.
- [ ] **Clean Lint & Types:** Typecheck and linters pass with zero warnings.
DownloadWhy each line is there

Cursor or Copilot: which job goes where

I pick the tool for the job in front of me. This is the split I use on client work.

The jobI useWhy
A change that touches several filesCursorThe agent holds the plan, repo chat finds the call sites, and I get one multi-file diff to review.
First week on a client repoCursorRepo chat answers “where does auth happen?” faster than grep when the names are lying.
Teaching a team one habitCursorChat, agent, and project rules live in the same editor.
Local work in a file I knowCopilotInline complete keeps me in flow. Opening an agent for a CSS tweak is ceremony.
A team standardized on VS CodeCopilotI use what they have for inline work and keep Cursor for the slices I own.
Pairing on someone else’s laptopCopilotTheir editor, their setup. I do not install Cursor mid-call to rename a prop.

Disclosure: I am the Cursor Ambassador for Indonesia and a SpaceXAI Ambassador, and SpaceX owns Cursor. Weigh my picks with that in mind. Based on client work, written up on . Read the full comparison.

What Cursor will cost your team

Pick a team size and how hard people lean on agents. The estimate uses Cursor’s current plans and tells you the date I last checked them. For what each plan includes and why third-party models cost extra, read my Cursor pricing breakdown.

How does a typical person on the team work?

Suggested plan

Super $100 per person a month

15x Plus usage. The allowance refreshes weekly.

Monthly cost

$100

Prices exclude tax. Claude, GPT, and other third-party models are billed on top at API list price, so a team that lives on them pays more than this. Plus ($8): Grok models only.

Prices checked against cursor.com/pricing on . Plan suggestions follow Cursor’s own advice on that page.

What changed in Cursor lately

The releases that matter for shipping teams, with a link to each source.

  1. Origin, Cursor’s code host, opened an app marketplace with 20+ apps for deploys, CI, code review, security, and monitoring.

    Source
  2. Remote control: see and reply to agents running on your computer from the Cursor iOS app.

    Source
  3. Two review bots on Teams and Enterprise: Rollouts watches each change as it deploys, Security Review reports exploitable bugs on every PR.

    Source
  4. Projects (beta): a coordinator agent plans larger work, delegates it to subagents, and keeps shared context between them.

    Source
  5. Self-hosted machines: agents run tool calls on machines inside your own network.

    Source
  6. SpaceX completed its acquisition of Cursor. Cursor joined SpaceXAI.

    Source

Cursor questions I get most

How do I start using Cursor on an existing repo?+

Write half a page of project rules first: the stack, the folder map, the error shape, naming, and the test command. Then plan the change in chat with real file paths, run the agent on only those files, and review the diff like a person wrote it. The beginner tutorial on this page walks through it.

How much does Cursor cost in 2026?+

On October 14, 2026 Cursor had a free plan, then Plus at $8, Premium at $30, Super at $100, Ultra at $200 a month. Teams starts at $30 per seat. Third-party models such as Claude and GPT are billed on top at API list price. The estimator on this page works it out for your team size.

Is Cursor free?+

There is a free plan with limited agent and chat requests, which is enough to try it on a side project. Plus only runs Grok models. Premium is the plan Cursor itself recommends for people new to AI coding, and it adds third-party models at API list price.

Do Cursor project rules actually help?+

Yes, when they are short. Half a page that names the stack, the folders, and “do not create new packages” stops most invented architecture. A 2,000-line style guide gets half-read and drifts from the repo. The rules file I commit is on this page.

Should I use Cursor or Copilot?+

Split the job. I use Cursor for changes that touch several files and for repos I do not know yet, and Copilot for inline work in an editor the team already uses. The table on this page lists which job goes where.

Is Cursor part of SpaceX now?+

Yes. SpaceX completed the acquisition on August 14, 2026, and Cursor joined SpaceXAI. Third-party models are still available in Cursor, billed at API list price.

Who wrote this

I’m Aris Setiawan, the first Cursor Ambassador in Indonesia, with 13+ years of client work. The loop, the rules, and the tool picks on this page come from client repos I ship. Prices and release notes link to Cursor’s own pages.

What a Cursor Ambassador does

Want this loop running on your team’s repo?

I set up project rules and a review habit on one of your real repos, then help the rest of the team adopt it. English or Indonesian.