Papermark for Agents
Papermark for Agents is an agent integration layer that exposes Papermark secure data rooms through 43 typed MCP tools and a versioned REST API. AI clients like Claude, Cursor, and ChatGPT can create data rooms, upload documents, mint watermarked share links, read page-by-page analytics, and run grounded Q&A over files the token is allowed to see.
Most document-sharing tools were built for humans clicking through a UI. Papermark for Agents is aimed at teams that want fundraising, M&A, LP update, and sales proposal workflows driven by code or chat, with scoped pm_live_ bearer tokens, link-level permissions, dynamic watermarking, and an append-only audit log still enforced on every agent action.
Founders, M&A advisors, PE and VC IR teams, real estate brokers, and SaaS platforms embedding white-labeled rooms use it to provision per-investor or per-buyer data rooms, bulk-mint email-gated links, and pull dwell-time analytics back into Slack or a CRM without manual SharePoint duplication.
43 typed MCP tools map one-to-one to versioned REST endpoints for agent workflows
Scoped pm_live_ bearer tokens authenticate agents with revocable, permission-bound access
Agents create data rooms, upload documents, and mint NDA-gated links with dynamic watermarking
Page-by-page view analytics return dwell time, downloads, and per-visitor activity via get_view_analytics
Grounded search_documents Q&A cites answers back to allowed source pages only
Connect via npx @papermark/mcp-server locally or HTTP at mcp.papermark.com for browser clients
43 MCP tools cover the full data room lifecycle from provisioning through analytics and audit export
Scoped tokens, watermarking, and permissions apply to agent actions the same way they do in the UI
Works with Claude, Cursor, ChatGPT, custom HTTP agents, and the papermark CLI on one API surface
Open source core allows self-hosted agent workflows when compliance requires it
Page-by-page analytics let agents rank engaged investors or buyers without opening the dashboard
MCP and REST access require Business plan (€59 per month annually) or higher, above the free tier
Full data room automation features sit on Data Rooms plans starting at €99 per month annually
Agent setup still needs a Papermark account and manual token minting before the first MCP call
Pricing is listed in euros, which may add conversion overhead for US-only teams
What is Papermark for Agents?
Papermark for Agents exposes Papermark secure data rooms to AI clients through 43 MCP tools and a REST API. Agents can create rooms, upload files, mint watermarked links, read page analytics, and run grounded Q&A while scoped tokens respect the same permissions as the web product.
Which MCP clients work with Papermark for Agents?
Papermark for Agents supports any MCP client. Use stdio with npx @papermark/mcp-server for Claude Desktop and Cursor, or connect over HTTP at mcp.papermark.com for browser clients like Claude Connectors and ChatGPT Apps.
How do agents authenticate to Papermark?
Papermark for Agents uses scoped API tokens formatted as pm_live_ created at app.papermark.com/settings/tokens. Each token acts as a long-lived bearer credential and respects link-level and folder-level permissions on every tool call.
What can an agent do with Papermark data rooms?
Papermark for Agents lets an agent provision data rooms, upload and version documents, create password-protected links with email verification, read page-by-page analytics, manage visitor access, subscribe to webhooks, and export an append-only audit log.
Does Papermark for Agents have a free plan?
Yes. Papermark for Agents builds on Papermark, which includes a free plan with 50 documents and page-by-page analytics. Programmatic MCP and API access starts on the Business plan at €59 per month billed annually, and full data room agent workflows use the Data Rooms plan from €99 per month annually.
Is Papermark for Agents open source?
Yes. Papermark for Agents runs on Papermark, which is open source and self-hostable on GitHub. Managed self-hosting and enterprise support are available on the Custom plan for teams that need data residency inside their own infrastructure.

