LibreOffice MCP

An MCP server that lets AI assistants (Claude Code, Claude Desktop, and any other MCP client) create, read and edit documents with LibreOffice: text documents, spreadsheets and presentations — no GUI, no copy-paste.

Quick start

All you need is Docker. Three commands:

git clone https://github.com/harshithb3304/libre-office-mcp.git
cd libre-office-mcp
docker compose up -d --build          # builds the image and starts the server
claude mcp add --transport http libreoffice http://localhost:8000/mcp

That’s it. Every tool is enabled out of the box. The first build takes a few minutes (it downloads LibreOffice); after that the server starts in seconds. Your files live in the workspace/ folder next to docker-compose.yml (see The workspace folder).

Try it — one write, one read

Ask your AI:

Create notes.docx with the heading “Groceries” and a bullet list of milk, eggs and bread. Then read the file back and tell me what is in it.

It writes the document (writer_add_content), reads it back (writer_get_text), and workspace/notes.docx appears on your computer. Spreadsheets and slides work the same way: “Make budget.xlsx with 4 months of income and expenses and a chart”, or “Make a 3-slide deck about our Q3 plan and export it to PDF.”

Using another MCP client

Any client that accepts an MCP server URL works. In the standard MCP JSON:

{
  "mcpServers": {
    "libreoffice": { "url": "http://localhost:8000/mcp" }
  }
}

(Some clients also want "type": "http" next to the url. Where the file lives depends on the client — Claude Desktop, Cursor, Windsurf, Cline and others each document it. A few clients, e.g. VS Code, name the top-level key servers instead of mcpServers.)

Prefer no background server?

Let your MCP client start the container itself instead. Build the image once, then give the client this command (standard MCP JSON; on Windows double the backslashes in the folder path):

docker build -t libreoffice-mcp .

{
  "mcpServers": {
    "libreoffice": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "-v", "/path/to/my-workspace:/data", "libreoffice-mcp"]
    }
  }
}

For Claude Code: claude mcp add libreoffice -- docker run -i --rm -v /path/to/my-workspace:/data libreoffice-mcp. -i is required; do not add -t.

Everyday commands

The server listens on 127.0.0.1 only and has no authentication, so just your computer can reach it. Read docs/DOCKER.md before exposing it further.

How it works

• The Docker container is the MCP server: LibreOffice, its Python bridge and fonts are inside the image. Your AI app only needs to be told where to find it.
• The AI can only see one folder of yours — the workspace — which appears as /data inside the container.
• Every edit is done by LibreOffice itself and saved atomically.

---

The workspace folder

The container cannot see your computer; it only sees the folder you mount. We call it the workspace, and inside the container it appears as /data.

• Relative paths the AI uses (report.docx, clients/acme.docx) resolve to the workspace. It is not your terminal’s current directory unless you mount it (-v "$(pwd):/data").
• Absolute paths work only inside /data; anything else is refused. Paths from your own computer (C:\Users\…) don’t exist in the container, so just give file names.
• It lives on your disk, not in the image or container. Stopping or deleting the container does not touch it, and the image does not grow as you create documents.
• To work on an existing file, copy it into the workspace first.
• Close a file in other programs before the AI edits it. Some programs lock open files (the write is then refused); others keep showing the old content until you reload.

More detail and permissions on Linux hosts: docs/DOCKER.md.

---

Choosing which tools the AI gets

By default everything is enabled. More tools means more for the model to read, so you can narrow it down by adding flags after the image name:

Example: claude mcp add libreoffice -- docker run -i --rm -v /path/ws:/data libreoffice-mcp --apps writer. docker run --rm libreoffice-mcp tools --apps writer prints exactly which tools a set of flags exposes. Every flag also has a LIBREOFFICE_MCP_* environment variable. The office_* core tools (create, convert, info, batch edit…) stay available unless excluded.

---

What the AI can do

The full list with parameters is in docs/TOOLS.md (generated from the code).

Writer

• Read — text (paged), structure (every paragraph/table with an index), outline, find, paragraph formatting, styles, tables, images, comments, bookmarks, links, footnotes, page setup, headers/footers, tracked changes
• Write — paragraphs, headings, bullet/numbered lists, whole sections in one call (writer_add_content), page breaks, merge another file
• Format — bold/italic/colour/font/size on specific text, paragraph alignment & spacing, create/modify styles
• Tables — create, edit cells, add/delete rows & columns, merge cells, borders, shading, header rows, column widths
• Objects — images (embedded, resizable, alt text), hyperlinks, bookmarks, footnotes, review comments, table of contents
• Page — paper size, orientation, margins, header/footer text with “Page N of M”
• Review — turn change tracking on/off, accept or reject all changes

Calc

• Build — a whole table in one call (calc_write_table: header, rows, number formats, auto-width, filter), values and formulas (the usual =SUM(A1,B1) syntax works), fill/copy/sort/find & replace, insert/delete rows and columns, merge cells
• Format — fonts, colours, number formats, borders, column widths, conditional formatting and data bars, data validation
• Charts and objects — column/bar/line/pie/scatter/area charts, images, hyperlinks, cell comments
• Workbook — add/rename/move/copy/hide sheets, named ranges, freeze panes (xlsx), autofilter, CSV

Impress

• Build a deck in one call (impress_add_slides: title, bullets with levels, two columns, image, table, notes), or add/duplicate/move/delete slides one by one
• Shapes and text — text boxes, 37 shape types, lines and arrows, images, tables, fills/outlines/shadows, text formatting, hyperlinks, alignment and z-order
• Look and feel — whole-deck themes (impress_apply_theme), master styles, slide size, backgrounds, transitions, footers, speaker notes

Any document — create, info, copy, convert/export (PDF, docx, odt, html, txt…), and office_batch_edit (many edits, one load, one save, all-or-nothing).

Writer content is addressed by block index (0 = first paragraph or table): the AI calls writer_get_structure to see the indexes, then edits by index — no fragile text matching.

---

Safety

• Your file is never half-written. Each request edits a private copy and replaces your file atomically at the end. A crash, a timeout or a failing step in a batch leaves the original untouched.
• No races. Requests run one at a time, so two simultaneous edits to the same file never interleave (the second waits).
• Macros never run, even in documents that contain them.
• The container only sees the workspace (/data); paths outside it, .. and symlink escapes are refused.
• No lock files are created next to your documents, and a stuck or cancelled request is killed cleanly (no leftover LibreOffice processes).
• There is no undo. Edits are saved immediately; keep backups of important files (or use office_copy_document first).

---

Coming from v1?

v2 is a rewrite; tool names changed to be consistent (office_*, writer_*):

The old libreoffice:{path} MCP resource is not ported yet.

---

Documentation

License

MIT