01 / Describe
Natural-language app building
Describe the site or application you want in plain language. The AI development agent turns your requirement into an application plan, then into complete frontend and backend source code.
earthy-bronze
by
Nada Kh
Nada Kh
Comments (0)
Sign in to leave a comment
Project Tasks
141 tasks#1
Generate system requirement document
3m 3s0.05 cr used
Done#2
Generate personas & user flows
0m 7s0.05 cr used
Done#5
Architecture
0m 41s0.05 cr used
Done#6
Workspace task plan
2m 54s0.05 cr used
Done#7
Create flow for Builder User
0m 6sCredits in parent
Done#8
Create flow for Administrator
0m 6sCredits in parent
Done#9
landing page
5m 59sCredits in subtasks
Done#10
Login
2m 23sCredits in subtasks
Done#11
Sign Up
3m 20sCredits in subtasks
Done#12
User dashboard
8m 6sCredits in subtasks
Done#13
AI development workspace
11m 2sCredits in subtasks
Done#14
AI settings
7m 15sCredits in subtasks
Done#15
Admin dashboard
10m 34sCredits in subtasks
Done#16
landing page / Landing Header Slot
0m 0s0.85 cr used
Done#17
landing page / Landing Hero Prompt
0m 50s0.85 cr used
Done#18
landing page / Landing Product Demo
0m 56s0.85 cr used
Done#19
landing page / Landing Features
0m 58s0.85 cr used
Done#20
landing page / Landing How It Works
0m 49s0.85 cr used
Done#21
landing page / Landing Pricing
0m 48s0.85 cr used
Done#22
landing page / Landing Faq
0m 48s0.85 cr used
Done#23
landing page / Landing Footer Slot
0m 0s0.85 cr used
Done#24
App Header
0m 36sCredits in parent
Done#25
App Footer
0m 47sCredits in parent
Done#26
Fix landing page: Footer brand and context text overlap at tablet width.
1m 23sCredits unavailable
Done#27
Fix landing page: The hero’s primary Build My App action routes to sign-up, but the accepted visitor flow…
1m 54sCredits unavailable
Done#28
Fix landing page review findings: Footer brand and context text overlap at tablet width (+1 more)
0m 24s0.02 cr used
Done#29
Login / Header Slot
0m 0s0.85 cr used
Done#30
Login / Access Panel
0m 42s0.85 cr used
Done#31
Login / Footer Slot
0m 0s0.85 cr used
Done#32
Sign Up / Header Slot
0m 3s0.85 cr used
Done#33
Sign Up / Access Panel
0m 50s0.85 cr used
Done#34
Sign Up / Footer Slot
0m 1s0.85 cr used
Done#35
User dashboard / Dashboard Header Slot
0m 0s0.85 cr used
Done#36
User dashboard / Dashboard Project Masthead
0m 21s0.85 cr used
Done#37
User dashboard / Dashboard Project Collection
2m 7s0.85 cr used
Done#38
User dashboard / Dashboard Footer Slot
0m 0s0.85 cr used
Done#39
AI development workspace / Workspace Header Slot
0m 0s0.85 cr used
Done#40
AI development workspace / Workspace Project Context
0m 57s0.85 cr used
Done#41
AI development workspace / Workspace Chat Panel
1m 2s0.85 cr used
Done#42
AI development workspace / Workspace Preview Panel
0m 55s0.85 cr used
Done#43
AI development workspace / Workspace File Editor Panel
1m 27s0.85 cr used
Done#44
AI development workspace / Workspace Publish Band
0m 49s0.85 cr used
Done#45
AI development workspace / Workspace Footer Slot
0m 0s0.85 cr used
Done#46
AI settings / Header Slot
0m 0s0.85 cr used
Done#47
AI settings / Orientation Band
0m 35s0.85 cr used
Done#48
AI settings / Provider Registry
1m 11s0.85 cr used
Done#49
AI settings / Availability Monitor
0m 54s0.85 cr used
Done#50
AI settings / Footer Slot
0m 1s0.85 cr used
Done#51
Admin dashboard / Admin Header Slot
0m 0s0.85 cr used
Done#52
Admin dashboard / Admin Orientation Band
0m 47s0.85 cr used
Done#53
Admin dashboard / Admin Monitoring Grid
0m 31s0.85 cr used
Done#54
Admin dashboard / Admin Provider Health Band
0m 53s0.85 cr used
Done#55
Admin dashboard / Admin Oversight Tables
1m 6s0.85 cr used
Done#56
Admin dashboard / Admin Security Settings Band
0m 48s0.85 cr used
Done#57
Admin dashboard / Admin Footer Slot
0m 0s0.85 cr used
Done#58
Fix User dashboard: Mobile create-project input text is clipped and overlapped by its button; see selectors…
2m 13sCredits unavailable
Done#59
Fix User dashboard review finding: Mobile create-project input text is clipped and overlapped…
0m 56s0.02 cr used
Done#60
Fix invoices.js: Unresolved import in…
0m 7s0.02 cr used
Done#61
Fix AI development workspace: The live-preview status label is clipped on desktop and tablet…
2m 13sCredits unavailable
Done#62
Fix AI development workspace: The selected file name is clipped in the tablet explorer row…
3m 32sCredits unavailable
Error#63
Fix AI development workspace review findings: The live-preview status label is clipped on desktop and… (+1 more)
1m 0s0.02 cr used
Done#64
Fix Admin dashboard: Non-scrollable panels clip provider timestamps and security settings controls at…
2m 6sCredits unavailable
Done#65
Fix Admin dashboard review finding: Non-scrollable panels clip provider timestamps and…
0m 22s0.02 cr used
Done#66
Fix AI development workspace review finding: The selected file name is clipped in the tablet explorer…
1m 17s0.02 cr used
Error#67
Create backend_modules diagram
0m 20sCredits in parent
Done#68
Create er_diagram diagram
0m 20sCredits in parent
Done#69
Create frontend_modules diagram
0m 20sCredits in parent
Done#70
Create sequence diagram
0m 20sCredits in parent
Done#71
Create service_network diagram
0m 20sCredits in parent
Done#72
Admin console for providers, users, and platform monitoring
140 cr estimated
To Do#73
Durable projects with unlimited creation and owner-only export
140 cr estimated
To Do#74
Secure sign-up, sign-in, sessions, and account settings
140 cr estimated
To Do#75
AI generation of complete frontend and backend source code
160 cr estimated
To Do#76
Conversational iteration, publishing, deployments, and exports
140 cr estimated
To Do#77
Dashboard project list with open, duplicate, delete, and thumbnails
120 cr estimated
To Do#78
Deployment history and source-code ownership on export
100 cr estimated
To Do#79
AI provider and model configuration with availability checks
120 cr estimated
To Do#80
AI agent turns natural-language requests into application plans
140 cr estimated
To Do#81
Connect and route AI providers from the AI settings page
140 cr estimated
To Do#82
Run each project in an isolated sandbox with a dev server and terminal
160 cr estimated
To Do#83
Keep secrets out of generated apps and log sandboxed execution
140 cr estimated
To Do#84
Refresh the live preview after AI changes project files
100 cr estimated
To Do#85
Create, rename, duplicate, and export projects with confirmation
120 cr estimated
To Do#86
Administrator panel shows usage and model availability
100 cr estimated
To Do#87
Dashboard project listings with search, recent projects, and management
120 cr estimated
To Do#88
Save, rebuild, and preview project changes with version history and export
140 cr estimated
To Do#89
Configure AI providers and models, test connections, and stream responses
120 cr estimated
To Do#90
Generate, iterate, preview, and publish applications from conversation
140 cr estimated
To Do#91
Persist projects durably across sessions
80 cr estimated
To Do#92
Secure uploads, protect API endpoints, and validate input
100 cr estimated
To Do#93
Stream AI responses with provider fallback and error handling
120 cr estimated
To Do#94
Browse, edit, create, rename, and delete project files
100 cr estimated
To Do#95
Manage user profiles, account settings, and personal AI provider credentials
100 cr estimated
To Do#96
Stream AI chat, live preview, export, and publish in the workspace
160 cr estimated
To Do#97
AI edits real project files, fixes errors, and generates schemas
160 cr estimated
To Do#98
Secure accounts: register, sign in, reset password, and manage profile
140 cr estimated
To Do#99
Connect personal AI providers with fallback and health monitoring
140 cr estimated
To Do#100
Administrator AI provider settings and defaults
120 cr estimated
To Do#101
Administrators manage application settings
80 cr estimated
To Do#102
Refresh the live preview from current project files
80 cr estimated
To Do#103
Application authentication gate
100 cr estimated
To Do#104
Record and view project version history
120 cr estimated
To Do#105
Enforce per-user project authorization
140 cr estimated
To Do#106
Persist and read user records in Postgre SQL
100 cr estimated
To Do#107
Throttle excess requests with rate limiting
100 cr estimated
To Do#108
Route AI requests through an independent multi-provider gateway
140 cr estimated
To Do#109
applications from ideas with streaming AI
160 cr estimated
To Do#110
Serve cloud generation through Open AI-compatible APIs
100 cr estimated
To Do#111
Automatically save project changes
100 cr estimated
To Do#112
Serve AI generation through a local Ollama model
120 cr estimated
To Do#113
Store and retrieve project files durably
120 cr estimated
To Do#114
Block unauthorized access to private projects
120 cr estimated
To Do#115
Store and handle secrets securely
120 cr estimated
To Do#116
Let users run exported projects locally
100 cr estimated
To Do#117
Record and expose AI usage
120 cr estimated
To Do#118
Serve AI requests through v LLM cloud AI
120 cr estimated
To Do#119
Serve AI requests through Ollama local AI
120 cr estimated
To Do#120
Persist and retrieve user profile records
100 cr estimated
To Do#121
Route generation requests to a v LLM cloud provider
120 cr estimated
To Do#122
Enforce database access policies
120 cr estimated
To Do#123
Optimize platform runtime performance
100 cr estimated
To Do#124
Save AI-made project file changes durably
120 cr estimated
To Do#125
Enforce platform-wide security on operations
140 cr estimated
To Do#126
an application plan and complete source code from a request
160 cr estimated
To Do#127
Process queued background jobs
120 cr estimated
To Do#128
Enforce that users only reach their own private projects
120 cr estimated
To Do#129
Store and retrieve project files durably
100 cr estimated
To Do#130
Store and retrieve conversation records durably
100 cr estimated
To Do#131
Apply database migrations to bring the schema up to date
80 cr estimated
To Do#132
Let users reset a lost password
100 cr estimated
To Do#133
Wire the AI settings page to live provider configuration
140 cr estimated
To Do#134
Wire the AI development workspace to live build, preview, and publish
160 cr estimated
To Do#135
Wire the admin dashboard to live platform management
140 cr estimated
To Do#136
Make the dashboard project list live
140 cr estimated
To Do#137
Make the landing page prompt and motion real
100 cr estimated
To Do#138
Build the workspace left panel for AI requests
140 cr estimated
To Do#139
Build the administrator panel for settings and AI monitoring
120 cr estimated
To Do#140
Build authentication, sessions and account access
140 cr estimated
To Do#141
Build the live preview center panel
120 cr estimated
To Do#142
Build settings and account management
120 cr estimated
To Do#143
Build the file explorer and code editor right panel
120 cr estimated
To Do
AI website & app builder
Turn Your Ideas Into Real Applications With AI
Describe what you want to build. The AI development workspace turns your words into plans, files and working source code you own.
No account needed to explore. Your idea carries into sign-up and is kept for your first project. Already have an account? Sign in
Product demonstration
One workspace where ideas become applications.
Chat Build session
Your promptDescribe the app in natural language.
Agent plan
Message the agent…↑
Live preview · desktop
Files Editor
src/app/page.tsx
src/components/Hero.tsx
src/app/api/route.ts
1export default function Page() {2return (3<main>4<Hero title="Your idea" />5</main>6)7}
The three-panel AI development workspace: your conversation on the left, the live application preview in the centre, and the file explorer with code editor on the right — shown above as an illustrative mockup.
Chat · Files · PreviewWhat you can build with
Every capability is real, every file is yours.
earthy-bronze turns a sentence into a working application — planned, coded, previewed and exported with source code you own. These are the parts of the platform that make that happen.
02 / Workspace
Three-panel AI development workspace
Build in one professional workspace: conversations with the agent, your project files and a live application preview, separated by 12px clay gutters.
03 / Gateway
Multi-provider AI Gateway
An independent AI Gateway sends requests to the provider you choose. Configure Ollama or vLLM endpoints, or any OpenAI-compatible provider, with streaming responses, fallback and error handling.
04 / Preview
Live preview with device modes
Test AI-generated applications safely in real time across desktop, tablet and mobile preview modes, without executing untrusted generated code on the main application server.
05 / Code
File explorer and code editor
Browse every project file in the explorer and edit it in the code editor. Changes are saved automatically, and version history keeps your project information persistent.
06 / Ship
Export and publishing
Export complete projects and keep ownership of the source code, or publish applications at a unique project URL.
07 / Own it
Self-hosting and source ownership
Run the platform on your own hardware with local AI models and self-hosted infrastructure, with secure authentication, access control and isolated code execution for private projects.
How it works
From a sentence to a working application
Five moves along one continuous ribbon: you describe, the agent plans and writes real source code, and you keep refining in conversation until you export or publish.
- 01STEP
Describe your idea
Type what you want to build in plain natural language — a portfolio, a storefront, an internal tool. No boilerplate, no configuration screens.
- 02STEP
AI analyzes and plans
The agent reads your requirements and produces an application plan: the screens, the data it needs, and the order the work should happen in.
- 03STEP
AI generates source code
Complete frontend and backend source files are written into your project, with schemas and dependencies installed as the plan requires.
- 04STEP
Review and refine conversationally
Keep talking to the agent. Ask for changes, fix what needs fixing, and explain what you want next — it preserves the functionality already in place.
- 05STEP
Export or publish
Take the complete source code with you, or publish the application under its own project URL. Code ownership stays yours either way.
Pricing philosophy
Build freely. Own everything.
earthy-bronze runs on a free usage philosophy: no mandatory subscriptions, no credit balances, no paywalls for the things that matter.
No artificial limits
Create as many projects as your ideas demand — there is no daily generation cap wired into the product.
- Unlimited project creation
- No artificial daily generation limits
- Reasonable abuse protection and infrastructure resource limits apply
Free at the core
The platform is designed around a free usage philosophy. No mandatory subscription, no artificial credit balance, and no paywall in front of core features.
- No mandatory subscriptions
- No artificial credit balances
- No paywalls for core features
Your own stack
Bring local or cloud AI providers and keep ownership of the source code. No mandatory paid AI API is required to build.
- No mandatory paid AI APIs
- Ollama and vLLM providers supported
- Export complete projects with source ownership
Self-host instead. Run earthy-bronze on your own hardware with local AI models to minimise operating costs — you keep the full source code and the infrastructure.
Questions, answered
Describe a website or web application in natural language. The AI development agent turns that requirement into an application plan and complete frontend and backend source code, then keeps modifying the real project files as you continue the conversation.
AI website & app builder
Turn Your Ideas Into Real Applications With AI
Describe what you want to build. The AI development workspace turns your words into plans, files and working source code you own.
No account needed to explore. Your idea carries into sign-up and is kept for your first project. Already have an account? Sign in
Product demonstration
One workspace where ideas become applications.
Chat Build session
Your promptDescribe the app in natural language.
Agent plan
Message the agent…↑
Live preview · desktop
Files Editor
src/app/page.tsx
src/components/Hero.tsx
src/app/api/route.ts
1export default function Page() {2return (3<main>4<Hero title="Your idea" />5</main>6)7}
The three-panel AI development workspace: your conversation on the left, the live application preview in the centre, and the file explorer with code editor on the right — shown above as an illustrative mockup.
Chat · Files · PreviewWhat you can build with
Every capability is real, every file is yours.
earthy-bronze turns a sentence into a working application — planned, coded, previewed and exported with source code you own. These are the parts of the platform that make that happen.
01 / Describe
Natural-language app building
Describe the site or application you want in plain language. The AI development agent turns your requirement into an application plan, then into complete frontend and backend source code.
02 / Workspace
Three-panel AI development workspace
Build in one professional workspace: conversations with the agent, your project files and a live application preview, separated by 12px clay gutters.
03 / Gateway
Multi-provider AI Gateway
An independent AI Gateway sends requests to the provider you choose. Configure Ollama or vLLM endpoints, or any OpenAI-compatible provider, with streaming responses, fallback and error handling.
04 / Preview
Live preview with device modes
Test AI-generated applications safely in real time across desktop, tablet and mobile preview modes, without executing untrusted generated code on the main application server.
05 / Code
File explorer and code editor
Browse every project file in the explorer and edit it in the code editor. Changes are saved automatically, and version history keeps your project information persistent.
06 / Ship
Export and publishing
Export complete projects and keep ownership of the source code, or publish applications at a unique project URL.
07 / Own it
Self-hosting and source ownership
Run the platform on your own hardware with local AI models and self-hosted infrastructure, with secure authentication, access control and isolated code execution for private projects.
How it works
From a sentence to a working application
Five moves along one continuous ribbon: you describe, the agent plans and writes real source code, and you keep refining in conversation until you export or publish.
- 01STEP
Describe your idea
Type what you want to build in plain natural language — a portfolio, a storefront, an internal tool. No boilerplate, no configuration screens.
- 02STEP
AI analyzes and plans
The agent reads your requirements and produces an application plan: the screens, the data it needs, and the order the work should happen in.
- 03STEP
AI generates source code
Complete frontend and backend source files are written into your project, with schemas and dependencies installed as the plan requires.
- 04STEP
Review and refine conversationally
Keep talking to the agent. Ask for changes, fix what needs fixing, and explain what you want next — it preserves the functionality already in place.
- 05STEP
Export or publish
Take the complete source code with you, or publish the application under its own project URL. Code ownership stays yours either way.
Pricing philosophy
Build freely. Own everything.
earthy-bronze runs on a free usage philosophy: no mandatory subscriptions, no credit balances, no paywalls for the things that matter.
No artificial limits
Create as many projects as your ideas demand — there is no daily generation cap wired into the product.
- Unlimited project creation
- No artificial daily generation limits
- Reasonable abuse protection and infrastructure resource limits apply
Free at the core
The platform is designed around a free usage philosophy. No mandatory subscription, no artificial credit balance, and no paywall in front of core features.
- No mandatory subscriptions
- No artificial credit balances
- No paywalls for core features
Your own stack
Bring local or cloud AI providers and keep ownership of the source code. No mandatory paid AI API is required to build.
- No mandatory paid AI APIs
- Ollama and vLLM providers supported
- Export complete projects with source ownership
Self-host instead. Run earthy-bronze on your own hardware with local AI models to minimise operating costs — you keep the full source code and the infrastructure.
Questions, answered
Describe a website or web application in natural language. The AI development agent turns that requirement into an application plan and complete frontend and backend source code, then keeps modifying the real project files as you continue the conversation.
No comments yet. Be the first!