Custom Actions Reference

Pipeline provides custom actions via the pipeline object. These extend Figma's capabilities with user interaction, AI requests, file downloads, and more.

---

pipeline.getUserInput(prompt, defaultValue?)

Show a modal dialog and wait for user input.

Opens an input dialog with a text field. Execution pauses until the user submits or cancels.

Parameters

Parameter	Type	Required	Description
prompt	string	Yes	Text shown above the input field
defaultValue	string	No	Pre-filled value in the input field

Returns

Promise<string> — The user's input, or defaultValue if cancelled

Examples

Ask for prefix

const prefix = await pipeline.getUserInput("Enter prefix:", "icon-");

for (const node of figma.currentPage.selection) {
  node.name = prefix + node.name;
}

console.log(`Added prefix "${prefix}" to ${figma.currentPage.selection.length} layers`);

Multi-step input

const width = await pipeline.getUserInput("Frame width:", "1920");
const height = await pipeline.getUserInput("Frame height:", "1080");

const frame = figma.createFrame();
frame.resize(parseInt(width), parseInt(height));
frame.name = `Frame ${width}x${height}`;

---

pipeline.aiRequest(prompt)

Make an AI request from your code.

Sends a prompt to AI and returns the response. Useful for content generation, analysis, translations, and intelligent automation.

Parameters

Parameter	Type	Required	Description
prompt	string	Yes	The prompt to send to AI

Returns

Promise<string> — The AI response text

Examples

Generate placeholder text

const topic = await pipeline.getUserInput("Topic for placeholder text:", "Technology");
const text = await pipeline.aiRequest(`Write 2-3 sentences about ${topic} for a website hero section.`);

const textNode = figma.createText();
await figma.loadFontAsync({ family: "Inter", style: "Regular" });
textNode.characters = text;

Translate content

const targetLang = await pipeline.getUserInput("Translate to language:", "Spanish");

for (const node of figma.currentPage.selection) {
  if (node.type === 'TEXT') {
    const translated = await pipeline.aiRequest(
      `Translate to ${targetLang}: "${node.characters}"`
    );

    await figma.loadFontAsync(node.fontName as FontName);
    node.characters = translated;
  }
}

Analyze design

const colors = figma.currentPage.selection
  .filter(n => 'fills' in n)
  .flatMap(n => n.fills)
  .filter(f => f.type === 'SOLID')
  .map(f => f.color);

const analysis = await pipeline.aiRequest(
  `Analyze this color palette for accessibility and harmony: ${JSON.stringify(colors)}`
);

console.log(analysis);

---

pipeline.downloadFile(data, filename, mimeType?)

Download a file to the user's computer.

Triggers a browser download with the provided data. Works with any binary data — images, SVGs, JSON, CSV, etc.

Parameters

Parameter	Type	Required	Description
data	Uint8Array	Yes	Binary data to download
filename	string	Yes	Name for the downloaded file
mimeType	string	No	MIME type. Default: 'application/octet-stream'

Returns

Promise<void>

Common MIME Types

Type	MIME
SVG	'image/svg+xml'
PNG	'image/png'
JPG	'image/jpeg'
PDF	'application/pdf'
JSON	'application/json'
CSV	'text/csv'
Plain text	'text/plain'

Examples

Export as SVG

for (const node of figma.currentPage.selection) {
  if ('exportAsync' in node) {
    const svg = await node.exportAsync({ format: 'SVG' });
    await pipeline.downloadFile(svg, `${node.name}.svg`, 'image/svg+xml');
  }
}

Export as PNG with scale

const scale = await pipeline.getUserInput("Export scale (1-4):", "2");

for (const node of figma.currentPage.selection) {
  if ('exportAsync' in node) {
    const png = await node.exportAsync({
      format: 'PNG',
      constraint: { type: 'SCALE', value: parseInt(scale) }
    });
    await pipeline.downloadFile(png, `${node.name}@${scale}x.png`, 'image/png');
  }
}

Export data as JSON

const data = figma.currentPage.selection.map(node => ({
  name: node.name,
  type: node.type,
  x: node.x,
  y: node.y,
  width: 'width' in node ? node.width : null,
  height: 'height' in node ? node.height : null,
}));

const json = new TextEncoder().encode(JSON.stringify(data, null, 2));
await pipeline.downloadFile(json, 'selection-data.json', 'application/json');

Export as CSV

const rows = [['Name', 'Type', 'Width', 'Height']];

for (const node of figma.currentPage.selection) {
  rows.push([
    node.name,
    node.type,
    'width' in node ? String(node.width) : '',
    'height' in node ? String(node.height) : '',
  ]);
}

const csv = rows.map(row => row.join(',')).join('\n');
const data = new TextEncoder().encode(csv);
await pipeline.downloadFile(data, 'layers.csv', 'text/csv');

---

pipeline.sleep(ms)

Pause execution for a specified time.

Async delay that respects cancellation. If the user clicks Stop during sleep, execution terminates gracefully.

Parameters

Parameter	Type	Required	Description
ms	number	Yes	Milliseconds to wait

Returns

Promise<void>

Examples

Animation delay

for (let i = 0; i < 10; i++) {
  node.x += 10;
  console.log(`Step ${i + 1}`);
  await pipeline.sleep(100); // 100ms between steps
}

Rate limiting API calls

for (const node of figma.currentPage.selection) {
  if (node.type === 'TEXT') {
    const translated = await pipeline.aiRequest(`Translate: ${node.characters}`);
    node.characters = translated;
    await pipeline.sleep(500); // Wait 500ms between API calls
  }
}

---

Note: For complete Figma API documentation, see the Figma Plugin API Reference.