usuarios
Creates synthetic user profiles from research data to validate service designs through natural conversation.
README
🧑🎨 usuarios · Synthetic User Profiles for Service Design
Create research-backed user profiles that validate your designs across every sprint.
usuarios is an MCP server that turns your service design research (interviews, observations, field notes) into synthetic user profiles — rich, 12-dimension archetypes you can use to validate designs, align teams, and test ideas. All through natural conversation in Claude Desktop or Codex Desktop.
🚀 What your team says vs. what happens
| They say | The AI does |
|---|---|
| "Creá usuarios sintéticos de las entrevistas" | Analyzes your research, extracts patterns, generates full profiles |
| "Validá el onboarding contra María" | Tests your design against María's criteria, returns a report |
| "¿Cómo va el proyecto?" | Shows a dashboard with research → patterns → profiles → validations |
| "Refiná el perfil de Juan" | Updates the profile with new insights, versions it |
Zero technical knowledge needed. Your team just chats.
📦 Installation (2 minutes)
1. Install uv
# macOS / Linux
curl -LsSf https://astral.sh/uv/install.sh | sh
# Windows
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"
2. Configure your AI desktop app
Claude Desktop: Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"usuarios": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/Sebtiago/usuarios-mcp",
"usuarios-mcp"
]
}
}
}
Codex Desktop: Edit ~/.codex/config.toml:
[mcp_servers.usuarios]
command = "uvx"
args = [
"--from",
"git+https://github.com/Sebtiago/usuarios-mcp",
"usuarios-mcp"
]
3. Restart your app and start chatting
"Inicializá usuarios para este proyecto"
That's it. The server handles everything else.
🧬 What's inside a profile? (12 dimensions)
Based on This Is Service Design Doing, Mapping Experiences, and the Touchpoint Journal:
| Dimension | What it captures |
|---|---|
| 1. Identity | Name, archetype, real quotes from research |
| 2. Empathy Map | Sees, hears, thinks/feels, says/does |
| 3. Jobs-to-be-Done | When/I want/So I can (functional, emotional, social) |
| 4. Pain Points | Intensity, frequency, context, traceability |
| 5. Behaviors | Patterns, triggers, workarounds |
| 6. Mindset | Beliefs, tech literacy, change attitude |
| 7. Ecosystem | Current tools, key people in their network |
| 8. Scenarios | Real usage flows with emotional arcs |
| 9. Emotional Journey | Stage-by-stage emotion map |
| 10. Validation Criteria | Intent principles + testable questions |
| 11. Traceability | Direct/Inferred/Speculative %, all sources cited |
| 12. Metadata | Version, expiration (12 months), human validation |
Every profile is saved in both JSON (machine-readable) and Markdown (team-readable).
🔄 The flow
INVESTIGACIÓN → ANÁLISIS → PERFILES → VALIDACIÓN → EVOLUCIÓN
(research/) (patterns/) (profiles/) (validations/) (versioned)
The AI host orchestrates everything automatically. You never touch the tools directly.
📂 Project structure
After initialization, your project looks like this:
your-project/
└── .usuarios/
├── config.yaml # Project settings
├── research/ # Drop your interview files here (.md, .txt)
│ ├── entrevista-1.md
│ └── focus-group.md
├── patterns/ # Extracted patterns (auto-generated)
│ ├── patterns.json
│ └── patterns.md
├── profiles/ # Your synthetic users (auto-generated)
│ ├── maria-cuidadora.json
│ └── maria-cuidadora.md
└── validations/ # Design validation reports
└── 2026-06-22-onboarding.md
🛠️ Development
# Clone
git clone https://github.com/Sebtiago/usuarios-mcp.git
cd usuarios-mcp
# Install dependencies
uv sync
# Run locally
uv run python main.py
# Customize templates (optional)
# Create .usuarios/templates/analyze.md in your project
# to override the default analysis methodology
🔒 Privacy
- Runs locally. No cloud, no API keys, no data leaves your machine.
- Does not call LLM APIs. The AI host (Claude/GPT) processes everything with its existing model.
- Your research data stays in
.usuarios/in your project folder.
📚 Methodology
This tool implements the service design methodology from:
- This Is Service Design Doing — Stickdorn, Hormess, et al.
- Good Services — Louise Downe
- Mapping Experiences — Jim Kalbach
- Touchpoint: The Journal of Service Design
- Analysis-Synthesis Bridge Model for AI in design
📄 License
MIT © Santiago Sirias
Built for designers, by a designer. If this helps your team, ⭐ the repo.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.