Code 0x faster
Stay 100% in control.

The AI-native IDE. Agents that write, plan, and debug — with any model, across your entire codebase.Specialized agents write, plan, and search — across your entire codebase, with any model.

Join Waitlist Read the Docs

Explorer

Open Editors

Ecommerce

▸.creor

▸.next

▸api

▸ecommerce-web

▸src/

▸app/

▸components/

⚛CartSummary.tsxM

⚛Navbar.tsx

⚛ProductCard.tsx

▸hooks/

▸lib/

Tapi.ts

Tutils.tsM

{}package.json

◇.envU

ⓘREADME.md

Outline

Timeline

CartSummary.tsxM

utils.tspage.tsx

ecommerce-web›src›components›CartSummary.tsx

import { useCart } from "@/hooks/useCart"

import { formatPrice } from "@/lib/utils"

export function CartSummary() {

const { items } = useCart()

const subtotal = items.reduce(

(sum, item) => sum + item.price * item.qty, 0

)

const total = subtotal

return (

<h2 className="text-lg font-semibold">Order Summary</h2>

{items.map((item) => (

<span>{formatPrice(item.price)}</span>

</div>

))}

<span>Total</span>

<span>{formatPrice(total)}</span>

</div>

)

}

PROBLEMSOUTPUTDEBUG CONSOLETERMINALPORTS

> ecommerce-web@0.0.0 dev

> vite

VITE v7.3.1 ready in 232 ms

→ Local: http://localhost:5173/

→ press h + enter to show help

12:07:48 AM [vite] (client) hmr update /src/components/CartSummary.tsx

⌥Creor Agent: ecommerce

⌥zsh

New Chat×

+◷|▢×

What can I help you with?

ecommerce main

Explain codeFix a bugWrite tests

Ask anything... "Write unit tests for auth"

Build Claude Opus 4.6

main*+◯ 0△ 1↑⚠ 0 ✕ 0

Sathvik (1 month ago)Ln 24, Col 1Spaces: 2UTF-8LF{} TypeScript TSX

Powering development with 19+ LLM providers

Anthropic

OpenAI

Google

xAI

Mistral

AWS Bedrock

Azure

Groq

Cohere

Perplexity

Together AI

DeepInfra

Cerebras

OpenRouter

Fireworks

Vertex AI

Anthropic

OpenAI

Google

xAI

Mistral

AWS Bedrock

Azure

Groq

Cohere

Perplexity

Together AI

DeepInfra

Cerebras

OpenRouter

Fireworks

Vertex AI

Agents that specialize

Not one AI, but many. Each agent has its own permissions, tools, and purpose. Auto-routing picks the right one.

FIG 0.1

const app = express()

app.get("/api", handler)

app.get("/api/v2", authMiddleware, handler)

app.listen(3000)

Build Agent

Full-access agent that writes code, runs commands, and executes multi-file changes in isolated git branches.

FIG 0.2

## Implementation Plan

Refactor auth middleware

Add rate limiting

Update API tests

Plan Agent

Read-only agent that explores your codebase and writes implementation plans — never touches your code.

FIG 0.3

Explore Agent

Deep codebase researcher. Traces symbols, maps dependencies, searches across every file.

Bring your own key. Use it unlimited.

Sign in with your API key from any provider — or authenticate through Creor. No middleman, no usage caps from us. Your key, your quota, your bill. Switch providers anytime.

1.0 Bring Your Own Key

API KeyConnected

sk-ant-•••••••••••••••••••••••Anthropic

Choose your provider

AAnthropic

OOpenAI

GGoogle

GGroq

AzAzure

+Custom

Your own API key

Paste key, no limits from us. Pay your provider directly.

Creor Auth

No usage caps from Creor.Unlimited

Your codebase, fully understood

Repo Map builds a structural overview of your entire project — auto-injected into every session. Context settings let you tune what's included: git state, file trees, token budgets. The AI doesn't guess your architecture — it sees it.

2.0 Codebase Intelligence

Repo Map

▸ src/

▸ app/

layout.tsx

page.tsx

▸ components/

Navbar.tsx

CartSummary.tsx

▸ lib/

utils.ts

api.ts

Token Budget

Drag to adjust

Excluded

.gitignorebinarieslockfilessecretsnode_modules

Auto-injected

Git stateFile treeDependencies

Teach it your standards

Skills embed team knowledge as interconnected markdown graphs — coding patterns, architecture decisions, domain context. Rules enforce policies in every AI interaction. Define once, apply everywhere.

3.0 Skills & Rules

Skill Graph

$How does JWT verification work?

SKILL.mdROOT

Auth system

↓ resolved 3 modules

oauth-flow

jwt-tokens

matched

middleware

↓ traced dependencies

refresh-flow

verify

rate-limit

←JWT tokens verified via verifyJWT() in auth middleware

Skill Graph

$How does JWT verification work?

←JWT tokens are verified via verifyJWT(token, SECRET) in auth middleware...

Rules

✓

code-style.md

Enforce naming conventions, formatting, import order

always

✓

security-policy.md

Block secrets in code, enforce auth patterns

always

✓

api-guidelines.md

REST conventions, error response shapes

on-demand

✓

testing-patterns.md

Unit test structure, mock strategy, coverage

on-demand

2 always · 2 on-demandall rules loaded

Extend with anything

25+ built-in tools for file ops, search, LSP, and analysis. Connect external services through MCP — Notion, Gmail, GitHub, or any custom server. Add your own tools with a single TypeScript file.

4.0 Tools & MCP

Built-in Tools25+ tools

📄

read

Read files

✏️

edit

Edit files

📝

write

Create files

⌨️

bash

Run commands

🔍

grep

Search content

📂

glob

Find files

🔗

lsp

Go to definition

git

Version control

🌐

web

Fetch URLs

+16 more tools

MCP Servers3 connected

GitHub

PRs, issues, actions

connected

Notion

Pages, databases

connected

Slack

Messages, channels

connected

Linear

Issues, projects

disabled

+Add MCP Server

Smart terminal, only when you need it

Most commands run silently in the background. But when the agent detects a long-running process, a dev server, or a command that needs your input — it opens a real interactive terminal automatically. Output streams back to the agent so it always knows what happened.

5.0 Interactive Terminal

Creor Agent

Creor Agent Terminal

Complete control, always

Hooks intercept every lifecycle event — before tool execution, on session start, after failures. Pattern-based permissions gate access per tool, per file, per agent. Nothing happens without your say.

6.0 Hooks & Permissions

Settings

⊙Account

⌂General

◈Models

⬡Custom Agents

🛡Permissions

◉Display

▸Terminal

◫Context

⚡Skills

☰Rules

📁Directories

⚙Tools & MCP

▷Hooks

🔧Troubleshooting

Permissions

Control how Creor handles tool execution permissions.

Quick Presets

SmartRecommended

Auto

Ask All

Per-Action Permissions

◎

Read Files

Read, glob, grep, list, code search

Auto

✎

Edit Files

Edit, write, patch, multi-edit

Auto

▸

Execute Commands

Shell / terminal commands

Ask

⊕

Web Access

Web search and fetch

Auto

⊞

MCP Tools

External MCP server tools

Ask

⬡

Tasks & Agents

Sub-tasks, skills, agents

Auto

Project Rules

No Project Rules Yet

Click 'Always (Project)' on a permission prompt or use the button below to add rules.

+ Add Rule

Hooks

Configure shell scripts that run at specific points during AI agent execution.

›◔

Before Tool Execution

›✓

After Tool Execution

›◧

Shell Environment

›⚡

Before Command

›◻

On Chat Message

∨▷

Session Start

Runs when a session begins. Use for setup, authentication, or context injection.

echo "$(date): session.start FIRED" >> /tmp/hooks.log

timeout 30s

+ Add Hook

›◉

Session End

›⊘

Tool Failure

›◔

On Notification

›◧

Before Compaction

Review every change. Revert any moment.

Every edit is shown as a diff before it touches your code — accept, reject, or retry. Snapshot revert lets you roll back specific files to any prior state in the session. Nothing is permanent until you say so.

7.0 Snapshot & Diff

Snapshot Timeline

a3f22 min ago

Added auth middleware

b7e15 min ago

Refactored API routes

c9d412 min ago

Initial session

Diff Preview

src/middleware/auth.ts

+0-0

14export function authMiddleware(req, res, next) {

15 const token = req.headers.auth

15 const token = req.headers.authorization?.split(" ")[1]

16 if (!token) return res.status(401).json({ error: "Unauthorized" })

17 try {

18 const user = verify(token)

18 const user = await verifyJWT(token, SECRET)

19 req.user = { id: user.sub, role: user.role }

20 next()

RejectAccept

Get started in minutes

From download to shipping code — three steps.

Step 01

Download & Open

Get Creor for macOS, Windows, or Linux. Open any project — your codebase is automatically mapped and ready for AI.

Step 02

Choose Your Model

Bring your own API keys from Anthropic, OpenAI, Google — or subscribe to Creor Gateway for unified billing across 19+ providers. Switch models anytime.

Step 03

Agents Build, You Ship

Describe what you want. Agents plan the approach, write the code, run the tests. Review inline diffs, approve changes, iterate — all without leaving the editor.

Ship faster. Stay in control.

Join the waitlist. Be the first to try Creor.

Join Waitlist See Pricing

Code 0x fasterStay 100% in control.

Agents that specialize

Build Agent

Plan Agent

Explore Agent

Bring your own key. Use it unlimited.

Your codebase, fully understood

Teach it your standards

Extend with anything

Smart terminal, only when you need it

Complete control, always

Permissions

Hooks

Review every change. Revert any moment.

Get started in minutes

Download & Open

Choose Your Model

Agents Build, You Ship

Ship faster. Stay in control.

Code 0x faster
Stay 100% in control.