Skip to main content
When a workflow input is a file, the SDK uploads it as multipart/form-data (the same shape as curl -F). No base64, no payload doubling.

​
Browser

import { EigenpalClient } from '@eigenpal/sdk';

const fileInput = document.querySelector<HTMLInputElement>('input[type=file]')!;
const client = new EigenpalClient({ apiKey });

await client.run('workflows.extract-invoice', {
  contract_document: fileInput.files![0], // File from <input type="file">
});

​
Node, from disk

import { readFile } from 'node:fs/promises';

const buffer = await readFile('contract.pdf');
await client.run('workflows.extract-invoice', {
  contract_document: {
    content: buffer,
    filename: 'contract.pdf',
    mimeType: 'application/pdf',
  },
});

​
Node, from a Blob

const blob = new Blob([buffer], { type: 'application/pdf' });
await client.run('workflows.extract-invoice', { contract_document: blob });

​
Multiple files

await client.run('workflows.compare-versions', {
  original: file1,
  revised: file2,
  reference: file3,
});
Each file becomes a top-level form field. Mix files and scalar inputs freely; scalars ride along in a single _json text field automatically.

​
Nested files aren’t extracted

Only top-level file values become multipart fields. Files inside arrays or nested objects stay in the JSON sidecar and the server won’t see them as uploads: 
// DON'T — `documents` becomes a JSON array of `{}` objects, no upload.
await client.run('workflows.compare', { documents: [file1, file2] });

// DO — flatten to top-level keys, your workflow accepts them by name.
await client.run('workflows.compare', { document_0: file1, document_1: file2 });

​
Don’t base64 yourself

// Don't do this. Doubles the payload size and skips the optimised path.
await client.run('workflows.extract-invoice', {
  contract_document: btoa(buffer.toString('binary')),
});
The SDK picks multipart whenever it sees a File, Blob, or { content, filename, mimeType }. Plain strings pass through as scalar inputs.