StitchSAFE
Universal MCP Server for Google Stitch. Connect AI agents to your UI designs.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://www.npmjs.com/package/stitch-mcp) [](https://opensource.org/licenses/MIT) [](https://modelcontextprotocol.io)
Universal MCP Server for Google Stitch
Build amazing UI/UX designs instantly with AI. Works seamlessly with Cursor, Claude, Antigravity, and any MCP-compatible editor.
Created by Aakash Kargathara.
🚀 How It Works
✨ Why this is cool
- Zero Config: Just login once and it works everywhere.
- Universal: Works on Windows, Mac, and Linux.
- Free: Google Stitch API is free to use.
�️ Quick Setup (2 Minutes)
1. Prerequisite: Google Cloud
You need a Google Cloud project with the Stitch API enabled.
# Login to your Google Cloud gcloud auth login # Set your project (replace with your actual project ID) gcloud config set project YOUR_PROJECT_ID gcloud auth application-default set-quota-project YOUR_PROJECT_ID # Enable the magic gcloud beta services mcp enable stitch.googleapis.com
2. Install the Credentials
This gives stitch-mcp permission to talk to Google on your behalf.
gcloud auth application-default login
3. Add to Your AI Editor
Copy and paste this into your MCP config file:
For Claude Desktop: `
43e206834132OBSERVED · 2026-10-07Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.
claude mcp add stitch-mcp -- npx -y [email protected]
{
"mcpServers": {
"stitch-mcp": {
"command": "npx",
"args": [
"-y",
"[email protected]"
]
}
}
}Exposed tools (2)
2 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
fetch_screen_code | read | Retrieves the actual HTML/Code content of a screen. Use this when you need to SEE the code. |
fetch_screen_image | read | Retrieves the screenshot/preview image of a screen. |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- declared (4 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (1)
@modelcontextprotocol/sdk, node-fetch
Gates applied: no_behavioural_pass.
43e206834132full audit observations/trust-audit/mcp-server/kargatharaakash__stitch.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-07 | 43e206834132 | SAFE | B | 89 | first audit |
Questions
What is the Stitch MCP server?
Universal MCP Server for Google Stitch. Connect AI agents to your UI designs.
What tools does Stitch expose?
2 in total: 2 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is Stitch safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does Stitch need?
No credential environment variables were found in its source, so it appears to need none.
How does Stitch run?
It speaks stdio, so it runs as a local process your client starts. It is published on npm as stitch-mcp at 1.3.2.
How current is this page?
The grade is for one exact copy of the source (43e206834132), read on 2026-10-07. The repository is watched and re-audited when it changes.