How We Build

The Robotville framework that powers me and scales from idea to product

Scroll to explore
Files
Docs
Agents
Rules
Templates

Without structure, AI projects become chaos

Managing multiple AI-powered projects requires more than just folders and files. As complexity grows, chaos creeps in: inconsistent structures, forgotten conventions, documentation drift, and agent confusion. Without a systematic approach, what starts as organized quickly becomes unmaintainable.

The Three Principles

The Robotville methodology is built on three core principles that ensure scalability and maintainability:

Dependency Pyramid

Each layer only depends on layers above it — changes cascade predictably. Rules at the base never change; implementation layers adapt to your needs.

Iterative Lifecycle

Ideas flow through documented phases: Ideation → Commercial → Development. Start simple, add complexity gradually, pivot based on learnings.

Template Enforcement

Consistency across all projects via mandatory document structures. Agents understand your work, documentation stays organized, quality is built-in.

Information Hierarchy in Detail

The Robotville repository uses a dependency pyramid where each layer conforms to and depends on the layers below it. This ensures changes cascade predictably and complexity never overwhelms structure.

L1: Meta Standards
Universal rules and behavioral standards. Constitution, conventions, design rules, methodology, library.
L2: Templates
Document structure blueprints. Templates for product overviews, presentations, design files.
L3: Project Implementation
Project-specific documentation and content. readme.md, roadmap.md, docs/*.md files in projects.
L4: Agent Definitions
Agent prompt files. villagers/*.md files that define how agents operate.
L5: Operational Artifacts
Runtime artifacts and cycle contracts. docs/contracts/*.md, logs, outputs.

L1 defines the rules → Universal standards that never change. L2 provides structures → Templates you follow ensuring consistency. L3-L5 implement → Your actual work built on solid foundations.

Three-Phase Lifecycle

The Robotville methodology defines an iterative, agile approach to organizing projects — from initial idea validation through live product. Every project moves through three phases, with iteration loops allowing refinement, pivoting, or returning to earlier phases based on learnings.

1

Ideation and Prototyping

Focus: Validate ideas through documentation, from concept to first prototype

Key Deliverables: product_overview.md, design files, visual prototypes (HTML/CSS/JS)

2

Commercial Validation

Focus: Create sales/commercial materials to test market fit

Key Deliverables: Sales presentations, pitch decks, discovery materials

3

Development and Pivoting

Focus: Build, iterate, and pivot until finding product-market fit

Key Deliverables: Working implementation, measured results, feature documentation

The Complete Picture

A visual summary of the three-phase methodology showing how ideas flow from inception to production.

Robotville Three-Phase Methodology Infographic

The Villagers

Specialized AI agents handle different stages of the cycle. Each agent has a specific role and creates specific outputs.

lavoisier avatar

lavoisier

Design Research
Analyzes inspiration sources, extracts design rules + tokens, creates design.md and design-[artifact].json files
designer avatar

designer

HTML Creation
Creates HTML documents, enforces token completeness, validates via browser MCP tools, generates static HTML output
designer_judge avatar

designer_judge

Visual Validation
Validates HTML documents via browser MCP tools (snapshot, network, console checks)
housekeeper avatar

housekeeper

Project Sync
Creates/updates project management files (readme.md, map.md), organizes folder structure
dom_cobb avatar

dom_cobb

Prompt Architect
Structures vague ideas into high-quality prompts, agent definitions, and general-purpose templates following prompt engineering best practices

Ideation and Prototyping Cycle

The Ideation and Prototyping Cycle is the first documentation cycle that takes a project from initial idea through completion of first documentation deliverables. This cycle demonstrates how Robotville's agent-based workflow operates.

Mode Discovery

Orchestrator discovers available modes

Design Research

lavoisier extracts design tokens

HTML Creation

designer creates documents

Validation

designer_judge validates output

Project Sync

housekeeper syncs files

Critical enforcement: All steps are mandatory. HTML-first rule means all documents must complete in HTML before any PDF conversion. Browser validation is mandatory and cannot be skipped.

Why This Matters

Robotville's methodology provides concrete benefits that make it worth adopting:

Predictable Structure

Every project follows the same patterns. When you create a new project, you know exactly where documentation goes, which templates to use, and how agents will interact with it. This predictability eliminates decision fatigue and ensures consistency across all your work.

Scalable Documentation

Documentation grows with your project. The methodology defines clear phases: Ideation → Commercial → Development. You start with minimal structure and add complexity only when needed. Templates ensure everything stays organized, even as your project evolves.

Agent Integration

AI agents understand and maintain your work because they follow the same methodology. When you invoke an agent, it knows where to look, which templates to apply, and how to keep your project synchronized. Agents aren't just tools — they're part of the system.

Proven Results

This methodology has been validated through real implementations across multiple projects. The layered architecture prevents chaos, the iterative lifecycle ensures quality, and template enforcement maintains consistency — even as projects grow in complexity.

me was built using Robotville. See it in action:

Back to me