Typed MCP actions
Agents can create data rooms, folders, links, and documents through typed tool calls instead of manual UI steps.
Papermark
Papermark for agents lets AI agents and custom software create data rooms, upload documents, mint secure links, and read page-level analytics via MCP and REST.
AI-accessible document sharing and data-room automation
Papermark for agents exposes the Papermark document-sharing platform to AI agents and software workflows through MCP and REST. It is designed for tasks such as creating data rooms, uploading documents, minting secure links, and reading page-by-page analytics without requiring manual work in the web app.
The source describes a setup built around scoped `pm_live_...` tokens, typed MCP tools, and versioned REST endpoints. Agents can work with data rooms, documents, links, visitors, and analytics while staying within the permissions of the token they use, and the platform records actions in an append-only audit log.
What Papermark gives agents and developers
Versioned REST API
Every tool maps to a versioned REST endpoint, so custom agents and server-side workflows can use the same surface over HTTP.
Command-line automation
The developer page shows a `papermark` CLI with OAuth device flow or token-based login, plus JSON output for pipelines.
Link and access controls
Share links can include passwords, expiry dates, email gating, download controls, watermarking, and allow/block lists.
Visitor and analytics reporting
Agents can list visitors, inspect page-by-page engagement, and read view analytics to understand how documents are used.
Events and audit trail
Webhooks and an append-only audit log let automations respond to events and preserve a record of views, downloads, and changes.
Practical workflows with Papermark
Fundraising data-room setup
A founder can ask an agent to create a data room, upload a folder of diligence files, organize them into folders, and generate password-protected, email-verified links for investors.
M&A buyer rooms
An advisor or banker can provision a buyer-specific room after an NDA, upload transaction documents, apply firm-level watermarking, and send the link back without duplicating the whole workspace manually.
Investor reporting and LP outreach
An IR or partnership team can distribute quarterly materials through unique links, then review who opened them, which pages they spent time on, and where follow-up is needed.
Real-estate listing delivery
A broker can forward property PDFs into a monitored workflow that creates a room, files documents into preset folders, and returns a shareable listing page with gated access.
Embedded document sharing
A product or platform team can embed Papermark behind the scenes and provision rooms and links from its own backend while keeping its own user interface.
Pros and Cons
Pros
- Lets agents create and manage data rooms, documents, and links through typed tools or REST calls.
- Supports secure sharing controls such as passwords, expiry dates, email verification, watermarking, and permissions.
- Provides page-by-page engagement analytics and visitor history for follow-up or reporting.
- Includes webhooks and an append-only audit log for event-driven workflows and traceability.
- Offers the same underlying surface through MCP, the CLI, and the REST API.
Cons
- The source does not show full setup details for every client, so some integration specifics are still unclear from the page set alone.
- The pricing page indicates plan-based limits for team members, links, documents, and data rooms, so capacity depends on the selected plan.
FAQ
How do agents connect to Papermark?
Papermark for agents exposes Papermark through typed MCP tools and versioned REST endpoints. The developer page shows a quickstart that uses a scoped token, the CLI device flow, or the SDK to make the first API call.
What interfaces are available for automation?
The source shows MCP access through `@papermark/mcp-server`, REST access over HTTP, and CLI access through the `papermark` command. The same token can be used across these surfaces.
What can an agent do with Papermark?
The developer page shows actions such as creating data rooms, uploading documents, creating links, listing visitors, and reading page-level analytics. The agents page adds workflows for webhooks, audit logs, and grounded Q&A over allowed documents.
How does Papermark handle access and auditability?
Papermark uses scoped `pm_live_...` tokens and an append-only audit log. The product pages also describe password protection, link expiration, email verification, watermarking, and permissions as available controls.
Are there any setup or usage limits shown in the source?
The source does not list a hard limitation for team usage, but it does show that some plans include a fixed number of team members and that enterprise support and self-hosting are separate options on the pricing page.
Quick Facts
- Category
- Developer Tool
- Primary interfaces
- MCP server, REST API, CLI
- Primary users
- Agents, developers, founders, advisors, and deal teams
- Source domain
- papermark.com
- Notable controls
- Scoped tokens, secure links, audit log, webhooks
- Pricing signal
- Free and paid plans are shown on the pricing page
Papermark Alternatives
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repos and verifies changes with compilers, debuggers, sanitizers, and tests.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads in Firecracker micro-VMs with private networking and SDK, CLI, or MCP control.
hob
hob is an independent workspace for coding agents, with local control over sessions, terminals, history, routing, and follow-up work.
Manta AI
Manta AI is an autonomous web app testing tool that maps app behavior, catches regressions, and generates tests from a URL, no scripts or selectors needed.
SonOf
SonOf connects to your repo and PM tool, audits your codebase, and turns approved work into shipped tickets with senior engineering review.
Ghost
Ghost is a terminal-based AI assistant for chatting, code generation, and CLI tasks. Includes free models, supports Linux, macOS, Windows, and is open source.