GroupDocs.Annotation MCP Server Leave feedback

GroupDocs.Annotation MCP server lets AI agents like Claude, Cursor, and Copilot annotate and review documents — PDF, Word, Excel, PowerPoint, images, and 50+ more formats — locally on your machine. Highlights, comments, arrows, strikeouts, and full reply threads, with nothing uploaded to any cloud service.

Run it with one command. The Docker image is self-contained — the runtime and every native dependency the engine needs are inside it:

docker run --rm -i -v $(pwd)/documents:/data \
  ghcr.io/groupdocs-annotation/annotation-net-mcp:latest

With the .NET 10 SDK installed, the same server also runs without Docker:

dnx GroupDocs.Annotation.Mcp --yes

Both are the .NET build of the server and run on Windows, Linux, and macOS. Other platforms will each get their own launcher — see Install for your platform.

Or use the guided installer to register the server in your AI client, verify the setup, and configure shared folders in one pass.

Eleven tools, which fall into four groups (full details in the tools reference):

Read — get_annotations lists every annotation as JSON (ids, types, pages, authors, replies); get_document_info returns pages and dimensions.

Write — add_annotation adds a highlight, comment, arrow, area, point, underline, or strikeout; update_annotation edits one; remove_annotations deletes by id or clears all.

Discuss — add_reply and remove_replies turn annotations into threads, with authors.

Move and show — export_annotations / import_annotations carry a round of comments to a new draft; generate_pages_preview renders pages as images the client displays inline.

Ask in plain language — “highlight the payment clause and ask legal to confirm”, “which comments are still unanswered?” — and the agent picks the tools.

Installation, prerequisites, and client configuration are platform-specific; the tools and licensing model below are the same everywhere.

Most document servers are write-only: the agent calls a tool and trusts the result. generate_pages_preview returns rendered PNGs inline, annotations included, so in a client with image support the agent can look at the page it just marked up, check the placement, and fix it before you ever open the file. Details: Let the agent see the page.

Exact config blocks for every client: Register in AI clients.

The server uses MCP’s local stdio transport: your AI client starts the server as a child process and talks to it over standard input/output. No inbound ports, no external endpoints, no telemetry — the data path is agent → local server → local filesystem. Review comments are often more sensitive than the document itself; this keeps both on your machine. Details: On-premise architecture.

A PDF reader can annotate a PDF. Choose this server when you need: the same annotation model across 50+ formats (Word, Excel, PowerPoint, images — not just PDF); annotations as structured data an agent can filter, count, and report on; threads with authors rather than loose notes; export/import so a review survives a new draft; and the fidelity of the commercial GroupDocs engine trusted by enterprise teams for over a decade.

• Quick start · Use cases · Troubleshooting & FAQ
• GitHub: server source · installer · integration tests
• NuGet package · Docker image · MCP Registry
• Questions: Annotation forum