Papely

By Papely

Get New App Alerts

Open in ChatGPT

Open in Claude

Overview

Papely turns a PDF you already have into a reusable template with editable fields (text, dates, images, QR codes and tables) and generates personalized documents from your data. The Papely MCP server brings the same capability to AI assistants that support the Model Context Protocol, so you can create documents by asking in plain language. An assistant connected to Papely can list the templates in your workspace and read the fields each one accepts; generate a single PDF, or up to 100 in one batch that downloads as a ZIP; check when a generation is ready and get a time-limited download link; register an image from a public URL to use in an image field; and email each document to its recipient, then see whether it was delivered and downloaded. You connect by signing in to Papely and approving the assistant on a consent screen, and you can revoke that access at any time. The assistant only reaches the templates, documents and images of your own workspace. Each generated document counts toward your workspace's generation quota, exactly as if you had generated it in the Papely web app.

Tools

generate_batch

ChatGPT

Generates 1–100 PDFs from one template as a single asynchronous job (one generationId; the download is a ZIP). Each entry of documents is a data object exactly as in generate_document. To email the PDFs, pass the shared email template block and put each recipient in that document’s data.email — documents without an email are generated but not sent. Each document consumes one generation of the workspace quota, charged atomically.

generate_document

ChatGPT

Generates one PDF from a template, asynchronously: returns a generationId to poll with get_generation_status and download with get_generation_document. data is keyed by field key from get_template; every declared field is optional (absent renders blank/default). Value shapes: text/date/qr = string, image = "" or an img_<id> from register_image_from_url, table = array of positional rows (one string-or-null cell per column, in columns[] order). To email the PDF, put the recipient in data.email (and optionally data.name) and pass the email template block. Each call consumes one generation of the workspace quota.

get_generation_document

ChatGPT

Returns a fresh, time-limited download URL for a COMPLETED generation (a single PDF, or a ZIP for a batch). The URL expires after a short window — fetch it right before downloading, and call this tool again if it expired. A generation still pending/processing reports GENERATION_NOT_FOUND; poll get_generation_status first.

get_generation_emails

ChatGPT

Returns the email-delivery status of a generation that opted into email. One entry per recipient: status pending/sent/delivered/bounced/failed plus ISO timestamps; a non-null downloadedAt means the recipient opened the download link. emailEnabled:false means email was never requested; emailEnabled:true with an empty list means it is still generating.

get_generation_status

ChatGPT

Returns a generation’s lifecycle status: pending → processing → completed (then download with get_generation_document) or failed (with an error message). Poll every few seconds until terminal.

get_template

ChatGPT

Returns one template with its declared fields. Field types: text/date/qr take a string, image takes "" or an img_<id> reference from register_image_from_url, and table takes an array of rows where each row is a positional array with exactly one string (or null) cell per entry of the field’s columns[], in order. Fields flagged system:true are the optional delivery columns email/name that go INSIDE data when sending the document by email.

list_templates

ChatGPT

Lists the PDF templates of this workspace, newest first. Use get_template to discover the fields a template accepts before generating a document from it.

register_image_from_url

ChatGPT

Downloads a PNG or JPEG from a public https URL and registers it for this workspace, returning an img_<id> reference to use as the value of an image field in generate_document/generate_batch. The image is stored once and can be reused across generations. Counts against the workspace storage quota.