Standardization

Shared resource

Start every project from a proven foundation. Your AI assistant uses shared templates to set up new projects and add features that follow your established patterns.

01 — Why Templates Matter

When your AI assistant creates something from scratch, it makes reasonable guesses about structure, naming, and conventions. Those guesses are different every time. After a few projects, you end up with a mix of styles that makes everything harder to maintain.

Templates solve this by giving your assistant a clear starting point. Instead of guessing, it generates projects and features from patterns your team has already approved. The result is consistent, predictable output — every time.

02 — What Your Assistant Can Do

New projects

Start from a template

Your assistant browses available project templates, picks the right one for the job, and sets everything up with the correct structure, configuration, and naming — all in one step.

&check;Browse and select from available templates
&check;Customize project name and settings
&check;Generate all starter files at once

Existing projects

Add features consistently

When your project needs new pages, components, or modules, your assistant adds them using the same templates that built the original project. New code fits in seamlessly.

&check;Discover features available for your project type
&check;Generate code that matches existing patterns
&check;Automatically handle imports and references

03 — What It Looks Like

Here’s what happens when you ask your assistant to create a new project from a template. It checks what’s available, sets up the foundation, and reports back:

You say

“Create a new web app called client-portal using our standard React template.”

Your assistant does

1.Checks available templates and finds the React starter
2.Generates the project with correct structure, routing, and configuration
3.Sets the project name, paths, and package references
4.Reports: “Created client-portal with 12 files from the react-spa template”

Generated project structure

apps/client-portal/
├── package.json          # Dependencies and scripts
├── tsconfig.json         # TypeScript configuration
├── vite.config.ts        # Build configuration
└── src/
    ├── main.tsx          # Application entry point
    ├── router.tsx        # Route definitions
    ├── routes/
    │   └── index.tsx     # Home page
    └── styles/
        └── global.css    # Base styles

04 — Getting Started

Add to your AI assistant

Connect standardization tools to your AI assistant so it can create projects and features from your shared templates:

{
  "mcpServers": {
    "standardizer": {
      "command": "npx",
      "args": [
        "-y",
        "@agiflowai/scaffold-mcp",
        "mcp-serve",
        "--admin-enable"
      ]
    }
  }
}

Or run directly

You can also use these tools from your terminal without connecting to an AI assistant:

npx -y @agiflowai/scaffold-mcp mcp-serve --admin-enable

05 — Better Together

Standardization tools pair with architecture tools. Templates create a strong starting point, and architecture tools make sure everything your assistant builds afterward stays consistent.

Templates set the standard

Every project and feature starts from an approved structure. Your assistant doesn’t guess — it follows the blueprint.

Quality checks keep it consistent

After building, architecture tools verify the result against your project’s quality rules and flag anything that doesn’t match.

06 — Related

Architecture

Enforce design patterns and validate results with automated quality checks.

View Source

Explore the open-source toolkit, contribute, or customize templates for your project.