stitch::upload-to-stitch

To upload large design assets to Stitch without token limits, use a direct HTTP-based Python script to stream files to the Stitch API. This bypasses standard MCP base64 encoding limits, ensuring large files transmit accurately without truncation.

Supported file formats for uploading design documentation to Stitch include images (PNG, JPG, WebP), HTML, and Markdown files. The upload script handles these formats automatically during the batch creation request.

Yes, you need a valid Stitch API key and an active project ID to authenticate and execute the batch creation request. These credentials are required to securely upload local images, HTML, and Markdown files to your project.

Uploading UI mockups via standard MCP base64 encoding fails because complex visual assets exceed the AI model's output token limitations, causing truncation. Direct HTTP upload prevents this by streaming the files directly to the Stitch API.

The best way to sync complex design specifications to a Stitch project is using a Python script that streams local assets directly to the API. This facilitates reliable batch creation for large files that exceed standard encoding limits.