Stitch mcp create project

Skill gabelul/stitch-kit/skills/stitch-mcp-create-project

Creates a new Stitch project container (a design workspace). Call this when starting a new design session to get a projectId for screen generation.From its SKILL.md

Install

npx -y skills add gabelul/stitch-kit --skill stitch-mcp-create-project

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

3.2 KB, 784 tokens by cl100k_base, as published. Nobody here has run it

Stitch MCP — Create Project

Creates a new Stitch project. A project is a container that holds one or more generated screens. You need a projectId before you can call generate_screen_from_text.

Critical prerequisite

Only use this skill when the user explicitly mentions "Stitch" in their request. Never trigger Stitch operations silently during regular conversation.

When to use

Starting a new design task or app concept
The user says "create a Stitch project", "start a new project in Stitch", "design X using Stitch"
You need a projectId before calling generate_screen_from_text
Starting work in a new domain (keep related screens in one project)

Step 1: Extract a meaningful title

Infer a descriptive project title from the user's request:

User says	→ title
"Design a pet adoption app"	`"PawPals"` or `"Pet Adoption App"`
"Make a login page"	`"Login Page"`
"SaaS dashboard for analytics"	`"Analytics Dashboard"`
"E-commerce product listing"	`"Product Listing"`

If no clear name is evident, use a descriptive label: "Design Project [date]".

Step 2: Call the MCP tool

{
  "name": "create_project",
  "arguments": {
    "title": "[Descriptive project title]"
  }
}

Step 3: Extract and store the project ID — CRITICAL

The tool returns a name field in the format projects/NUMERIC_ID. You must extract the numeric portion:

Tool response:

{
  "name": "projects/3780309359108792857",
  "title": "Analytics Dashboard",
  "createTime": "2026-03-24T14:30:00Z",
  "visibility": "PRIVATE",
  "deviceType": "DESKTOP",
  "projectType": "TEXT_TO_UI",
  "designTheme": {
    "colorMode": "LIGHT",
    "customColor": "",
    "colorVariant": "TONAL_SPOT",
    "roundness": "ROUND_TWELVE",
    "spacingScale": 1,
    "headlineFont": "INTER",
    "bodyFont": "INTER",
    "labelFont": "INTER",
    "namedColors": {},
    "designMd": ""
  }
}

The designTheme populates with full values (including designMd and namedColors) after the first screen is generated. For a brand-new project, expect mostly empty/default theme values.

Extract:

Full name: projects/3780309359108792857 → use for list_screens, get_project
Numeric ID: 3780309359108792857 → use for generate_screen_from_text, get_screen

Store both values. Announce to the user:

"Created project 'Analytics Dashboard' (ID: 3780309359108792857). Ready to generate screens."

ID format rules (critical — different tools need different formats)

Tool	ID format required
`generate_screen_from_text`	Numeric only: `3780309359108792857`
`get_screen`	Numeric only: `3780309359108792857`
`list_screens`	Full path: `projects/3780309359108792857`
`get_project`	Full path: `projects/3780309359108792857`

Integration

After creating a project, the next step is always stitch-mcp-generate-screen-from-text to generate the first screen.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most mcp tooling skills give in 784 tokens

Counted across 780 of the 1,136 authors here whose files we hold, read 2026-09-06

Use Zod for input validationin 34 of 780, across 21 files
Use stdio for local clientsin 27 of 780, across 10 files
Restart Claude Code after configurationin 26 of 780, across 23 files
Verify MCP server connection before using toolsin 23 of 780, across 17 files
Define input schemas for every toolin 20 of 780, across 11 files
Use Streamable HTTP for remote clientsin 18 of 780, across 8 files
Pin SDK version in package.jsonin 17 of 780, across 6 files
Keep server logic independent of transportin 16 of 780, across 6 files
Verify SDK methods against official documentationin 15 of 780, across 5 files
Format evaluation results as an XML filein 15 of 780, across 12 files
Test servers using the MCP Inspectorin 15 of 780, across 14 files
Create ten complex and independent evaluation questionsin 14 of 780, across 11 files

Said here and by no other author read

Infer a descriptive project title from user request
Use a descriptive label if no title is evident
Call the create_project tool
Extract the numeric portion from the project name
Store both full name and numeric ID
Announce project creation and ID to the user

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.