Papermark is the secure data room built for modern dealmakers. Share confidential documents on a platform that gives you control, visibility, and trust. Build and manage your data room for buy-side deals or raising funds. Papermark's data room can be fully whitelabeled with custom domains, built-in analytics and advanced security controls. Papermark is open-source with hosted and self-hosted options.
This is the 3rd launch from Papermark. View more
Papermark Agents
Launched this week
Papermark for Agents lets AI run your secure data rooms for you. Just ask: build a data room, upload and organize files, create watermarked links, or see which investors actually read your deck. 40+ MCP tools, plus a REST API and CLI. Open source, secure by default, and built for Claude, ChatGPT, Cursor, and your own code.
Free Options
Launch Team / Built With
Papermark
Hey Product Hunt 👋
Marc here, founder of Papermark.
We started as an open-source DocSend alternative. Fast forward to today, Papermark serves 60k+ companies with over $23B in deals managed through it.
Today we're shipping Papermark Agents to give AI agents complete control over your data room.
It's an MCP server, API, and CLI that let your agent do everything you'd normally do in the dashboard. Upload documents, create tracked access links, spin up data rooms, and pull visitor analytics, all from Claude, ChatGPT, and any other agent.
A few things you can now do just by asking your agent:
"Upload this deck and give me a tracked share link" → done in one message
"Create a data room for this deal and add these files"
"Who viewed my pitch deck this week?" → analytics back in chat
Open source at the core. Agent-native at the edges.
We think every data room should work with your agent, and we're the first to make that real.
What do you hand off to your agent first? 🚀
PageAI
@mfts0 this is insane, I LOVE IT. I didn't think Papermark could get any more awesome!
Fantastic release 👏
@mfts0 and @iuliia_sh , huge congratulations on shipping this! Manually configuring secure data rooms and minting watermarked links for every complex enterprise bid or client onboarding is a massive time sink. Exposing this entirely through an MCP server so an agent can handle the infrastructure and organization is a brilliant way to eliminate that friction.
Quick question on the architecture: Does the MCP server allow the local agent (like Cursor or Claude) to actually read and synthesize the contents of the documents inside the room, or is the integration strictly focused on managing the links, analytics, and folder infrastructure?
Papermark
@iuliia_sh @varunvivek thanks for support. currently it's focused on the managing aspect. we have another launch coming where we also give agents access to the dataroom's content.
It would work with local agents as long as they support the MCP protocol
@iuliia_sh @mfts0 Makes perfect sense! The infrastructure management alone is a fantastic use of the MCP protocol. Really looking forward to that next launch - Best Wishes.
Papermark
Hi awesome people👋
Every day we talk to customers who run their entire deal flow through Papermark and almost all of them are doing the same repetitive setup work by hand. Creating data rooms, organizing folders, minting links, checking who opened what. We kept thinking: an agent could do this. So we built the surface to make it real.
Today we're launching the thing we're most excited about: Papermark for Agents. 🤖
Point Claude, Cursor, or ChatGPT at Papermark and it actually does the work for you.
Connect it in one line: npx -y @papermark/mcp-server, drop in a token, and start typing in plain English.
Here's what Papermark power users already doing:
🚀 Founders raising a round - "Create a Series B data room, upload everything in ~/raise/, organize into Financials, Legal and Product, and mint a watermarked link for each VC on my list." What used to take 2–3 hours of setup happens in one prompt, while you grab a coffee.
💼 M&A advisors and investment bankers - The moment an NDA is signed, an agent spins up a per-buyer room, uploads the CIM and model, watermarks it with the buyer's firm name, and emails the link back. The associate work that used to take 30–60 minutes per buyer happens in seconds.
📊 Founders and sales teams chasing the right people - Ask "Which investors spent more than 60 seconds on the financials page?" and get a ranked list back. No more guessing who to follow up with — your agent reads the analytics for you.
🏦 VC and PE firms running LP communications - Send email-gated LP updates in bulk from a CSV, then two weeks later ask your agent to flag everyone who never opened it and draft personalized nudges.
🏠 Real estate brokers - Forward a listing email with the property PDFs to an inbox, and the agent builds the room, files the docs into the right folders, and replies with a ready-to-share link and QR code for the listing page.
🔌 SaaS platforms embedding Papermark — Provision white-labeled rooms server-side on your own domain via the REST API. Your users get the secure viewer, watermarking, and analytics — they never even see Papermark.
We would really love to hear what you think ☺️
Papermark being open-source and self-hostable already made it stand out in the data-room space. Adding agents that can "run a deal" is a bold step, interesting and a little nerve-wracking given how sensitive fundraise and buy-side docs are. How do you scope what an agent is allowed to do inside a room, and where does the human stay in the loop on access decisions? Permission boundaries feel like the whole ballgame here.
The MCP surface for data rooms is a smart move. In a regulated deal flow the audit trail matters as much as the action. When an agent spins up a room or mints a watermarked link, does the log capture that it was the agent acting on a person's behalf, and on whose authority? If something leaks you need to reconstruct who actually authorised the share, not just that a link was created.
The document chase in a B2B deal or fundraise is almost always where momentum stalls. Someone shares a data room, the other side goes quiet for three days, then comes back with a list of what is still missing. An agent that tracks what has been uploaded, reviewed, and flagged incomplete feels like it could cut that dead time significantly.
I am curious whether Papermark Agents can detect gaps proactively. Like if a deal partner expects to see specific document types and they are missing from the room, can the agent flag that before the review meeting rather than after it?
Giving an agent 40+ MCP tools over a data room is powerful, but 'create a tracked external share link to this confidential deck' is exactly the action I'd want gated. Can an agent token be scoped to specific data rooms or to read-only/analytics, so a misfired prompt can't mint an external-shareable link to the wrong document? And in self-hosted mode, do those MCP tools run fully against my own instance, or does link-tracking still route through Papermark's cloud?