PandaDoc MCP server for your AI

No. The connect form asks for one value, API Key, and nothing else — no OAuth app, no redirect URL, no client secret, and no re-authentication later. Generate the key in PandaDoc under Developer Dashboard → Configuration and paste it in. PopMCP authenticates over HTTPS with an Authorization: API-Key header, stores the key encrypted, and uses it server-side only.

Yes. PandaDoc's Configuration screen issues both Sandbox and Production keys, and the connection works against whichever one you paste. Point a connection at a sandbox key while you tune prompts and approval rules, then swap in a production key when you are happy. Either way the connection is scoped to that single account and reaches nothing outside it.

It can do both, if you let it. The write side covers creating a document from a template or from a PDF/DOCX URL, adding, editing and reassigning recipients, creating document fields, sending, changing status, moving documents into folders and firing manual reminders. Set the connection to read-only and only the lookups and downloads remain.

115 operations across 25 API modules, split 54 read and 61 write. The 42 most-used load into your AI client by default and the rest stay searchable, so a client with a tool limit is not flooded. Documents is the largest module, followed by user and workspace management, templates, folders and webhook subscriptions.

Document status changes between draft, sent, viewed, completed and declined; a recipient finishing their part; the signed PDF becoming available to download; documents being updated, deleted or failing to generate; sections added and quote line items changing; templates created, updated or deleted; and content library items created or failed. Those all arrive as webhooks. New drafts appear through a polled Document created trigger instead.

Creating a document and sending it are two separate calls, and a document is not sendable the instant it is created — PandaDoc generates it from the template first. That is why there is a dedicated document status call and a Document creation failed webhook for templates that never render. Have the agent check status before it sends, rather than chaining create straight into send.

No. PopMCP hosts it and gives you a Streamable HTTP MCP URL. There is nothing to deploy, no callback URL to register, and no infrastructure to keep running or patch.

Any client that supports a remote Streamable HTTP MCP server — Claude Desktop, Claude Web, ChatGPT custom connectors, Codex and the Codex CLI, OpenCode and Z.ai's ZCode, as well as editors such as Cursor, Windsurf and Zed. Your client signs in to PopMCP with OAuth; the provider credential never leaves the server.

Loading 115 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 42-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. You can switch a connection to the full catalog at any time.

A tool is something your AI calls — there are 115 here. A trigger is something PopMCP watches for you: 23 PandaDoc events that can start work on the PopMCP side. Triggers never enter your AI client's context, and none of them ask you to paste a callback URL into PandaDoc.