Dokmatiq DocGen
OfficialThis MCP server exposes Dokmatiq DocGen tools for AI assistants to generate, process, sign, inspect, and export documents and receipts.
Generate PDFs from HTML/Markdown and documents from templates with fields, watermarks, and PDF/DOCX/ODT output.
Compose multi-part documents and create ZUGFeRD-compliant invoice PDFs.
Work with PDFs: merge, extract text, read metadata, convert to PDF/A, rotate, preview pages, and count pages.
Inspect and fill PDF forms, optionally flattening fields.
Digitally sign PDFs with PKCS#12 certificates, verify signatures, and list uploaded certificates.
Manage templates and fonts: list/upload/delete templates and list custom fonts.
Generate, convert, fill, and inspect Excel workbooks (XLSX) and CSV data.
Validate, extract, parse, and detect e-invoices in ZUGFeRD/Factur-X and XRechnung formats.
Use AI to extract receipt/invoice data, run async jobs, and generate expense reports.
Export receipts as DATEV-compatible CSV or styled XLSX.
Enables exporting receipt data in DATEV-compatible CSV format for seamless integration with German accounting software.
Dokmatiq DocGen MCP Server
MCP (Model Context Protocol) server that exposes the Dokmatiq DocGen document generation API as tools for AI assistants like Codex, Claude Code, Claude Desktop, Cursor, Continue, Cline, and Hermes.
Use it to generate PDFs, DOCX/ODT documents, Excel workbooks, invoices, e-invoices (ZUGFeRD/XRechnung/Factur-X), receipt exports, signed PDFs, PDF forms, and documents on company letterhead (Briefpapier/Firmenpapier).
Setup
Prerequisites
Python 3.11+
DocGen API key
Installation
python3.11 -m pip install --user dokmatiq-docgen-mcpThe package requires Python 3.11 or newer. On macOS, the system or Xcode-provided python3 can be too old; use Homebrew Python, pyenv, uv, or another Python 3.11+ runtime if pip cannot find a compatible package.
pip install --user commonly installs the docgen-mcp command into ~/.local/bin. If your MCP client cannot find docgen-mcp, add ~/.local/bin to PATH or use the absolute command path in your MCP config.
Or install from source:
cd mcp
pip install -e .Running tests
From the repository root, use the checked-in lockfile to install the MCP server and run its stdio/client smoke tests:
uv run --project mcp python -m unittest discover -s mcp/tests -vConfiguration
Set your API key as an environment variable:
export DOCGEN_API_KEY=dk_live_xxxOptionally override the base URL:
export DOCGEN_BASE_URL=https://custom.api.endpoint.comRelated MCP server: rechnungsapi-mcp
Usage with Claude Desktop
Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"docgen": {
"command": "/Users/you/.local/bin/docgen-mcp",
"env": {
"DOCGEN_API_KEY": "dk_live_xxx"
}
}
}
}Or using uvx:
{
"mcpServers": {
"docgen": {
"command": "uvx",
"args": ["dokmatiq-docgen-mcp"],
"env": {
"DOCGEN_API_KEY": "dk_live_xxx"
}
}
}
}Usage with Claude Code
claude mcp add docgen -- docgen-mcpSet the API key in your environment or .env file.
Usage with Codex
Install the companion skill:
python3 ~/.codex/skills/.system/skill-installer/scripts/install-skill-from-github.py \
--repo dokmatiq/docgen-sdks \
--path plugin/skills/dokmatiq-docgenThen configure an MCP server named docgen using either docgen-mcp, an absolute path such as /Users/you/.local/bin/docgen-mcp, or uvx dokmatiq-docgen-mcp. Make sure DOCGEN_API_KEY is available to the server process.
Usage with Hermes
Add the MCP server to your Hermes config (~/.hermes/config.yaml):
mcp_servers:
docgen:
command: "uvx"
args: ["dokmatiq-docgen-mcp"]
env:
DOCGEN_API_KEY: "dk_live_xxx"Replace dk_live_xxx with your Dokmatiq API key.
Environment Variables in MCP Configs
Do not use shell placeholders such as ${DOCGEN_API_KEY} inside JSON MCP config files unless your specific client documents that it expands them. Many MCP clients pass that value literally to the server, which causes authentication to fail.
Available Tools
Document Generation
Tool | Description |
| Convert HTML to PDF |
| Convert Markdown to PDF |
| Generate with template, fields, watermark, output format |
| Compose multi-part document from sections |
| Create ZUGFeRD-compliant invoice PDF |
PDF Tools
Tool | Description |
| Merge multiple PDFs into one |
| Extract all text content |
| Get title, author, page count |
| Convert to PDF/A archival format |
| Rotate pages |
PDF Forms
Tool | Description |
| List form fields with types and values |
| Fill form fields |
Digital Signatures
Tool | Description |
| Digitally sign with PKCS#12 certificate |
| Verify existing signatures |
| List uploaded certificates |
Templates & Fonts
Tool | Description |
| List uploaded templates |
| Upload ODT/DOCX template |
| Delete a template |
| List uploaded fonts |
Preview
Tool | Description |
| Render page as PNG image |
| Get total page count |
Excel Workbooks
Tool | Description |
| Generate styled XLSX from structured JSON (sheets, columns, formulas, styling, freeze panes) |
| Convert CSV to formatted Excel with auto-filter and styled headers |
| Extract Excel sheet data as CSV |
| Extract Excel sheet data as structured JSON with typed values |
| Fill Excel template with values at named cells and ranges |
| Inspect workbook metadata (sheet names, row/column counts, named ranges) |
E-Invoicing
Tool | Description |
| Validate ZUGFeRD compliance |
| Extract invoice data from ZUGFeRD PDF |
| Validate XRechnung XML |
| Parse XRechnung into structured data |
| Detect XRechnung format |
Receipt Recognition (AI-Powered)
Tool | Description |
| Extract structured data from receipt image (vendor, totals, VAT, line items, SKR03/04 account) |
| Submit receipt for async extraction with optional webhook |
| Check async extraction job status |
| Get extraction result of completed job |
| List all async receipt jobs |
| Extract receipt and generate expense report (PDF/DOCX/ODT) |
| Export receipts as DATEV-compatible CSV |
| Export receipts as Excel workbook |
Example Prompts
Once connected, you can ask Claude:
"Generate a PDF from this HTML:
<h1>Hello World</h1>""Create an invoice for ACME GmbH to Kunde AG for 8 hours of consulting at 120€/hour"
"Merge these two PDFs together"
"Extract the text from this PDF"
"Fill the form fields name=Max and date=12.04.2026 in this PDF"
"Extract the data from this receipt image -- I need vendor, total, and VAT breakdown"
"Analyze these receipts and export them as a DATEV CSV"
"Turn this receipt photo into an expense report PDF"
License
MIT
Available Tools
40 toolscompose_documentA
Compose a multi-part document from several sections.
Each part is a dict with: htmlContent, markdownContent, templateName, fields.
Args: parts: List of document parts. Each part can have htmlContent, markdownContent, templateName, fields. watermark: Optional diagonal watermark text overlay. output_format: Output format – PDF, DOCX, or ODT (default: PDF).
Returns: Base64-encoded document bytes.
Inputs: parts (required), watermark, output_format. Returns: result
convert_to_pdfaB
Convert a PDF to PDF/A archival format.
Args: pdf_base64: Base64-encoded PDF file.
Returns: Base64-encoded PDF/A file.
Inputs: pdf_base64 (required). Returns: result
create_invoiceA
Create a ZUGFeRD-compliant invoice PDF with structured e-invoicing data.
Each item in the items list should have: description, unitPrice, and optionally quantity, unit (C62=piece, HUR=hour, DAY=day, KGM=kg), vatRate.
Args: invoice_number: Unique invoice number (e.g. "RE-2026-001"). invoice_date: Invoice date in ISO format (e.g. "2026-04-12"). seller_name: Seller company name. seller_street: Seller street address. seller_zip: Seller postal code. seller_city: Seller city. buyer_name: Buyer company name. buyer_street: Buyer street address. buyer_zip: Buyer postal code. buyer_city: Buyer city. items: Line items – each with description, unitPrice, and optionally quantity, unit, vatRate. iban: Payment IBAN. bic: Payment BIC/SWIFT code. payment_terms: Payment terms text. due_date: Payment due date (ISO format). seller_vat_id: Seller VAT ID. seller_country: Seller country code (default: DE). buyer_country: Buyer country code (default: DE). buyer_vat_id: Buyer VAT ID. currency: Currency code (default: EUR). template_name: Template name for the invoice layout. html_content: HTML content for the invoice body.
Returns: Base64-encoded PDF with embedded ZUGFeRD XML.
Inputs: invoice_number (required), invoice_date (required), seller_name (required), seller_street (required), +18 more. Returns: result
csv_to_excelA
Convert CSV content into a styled Excel workbook.
Auto-formats with header styling, auto-filter, frozen header row, and auto-sized columns.
Args: csv_content: Raw CSV text content. delimiter: CSV delimiter (default: comma). has_header: Whether the first row is a header (default: true). sheet_name: Name for the sheet tab (default: "Data").
Returns: Base64-encoded XLSX file.
Inputs: csv_content (required), delimiter, has_header, sheet_name. Returns: result
delete_templateB
Delete a template by name.
Args: name: Name of the template to delete.
Returns: Confirmation message.
Inputs: name (required). Returns: result
detect_xrechnungA
Detect if an XML document is an XRechnung and identify its format.
Args: xml: XML content to analyze.
Returns: JSON with detection result (detected, format, version).
Inputs: xml (required). Returns: result
excel_to_csvA
Extract data from an Excel sheet and return it as CSV text.
Args: excel_base64: Base64-encoded Excel file. sheet_index: Sheet index to extract (0-based, default: 0). delimiter: CSV delimiter (default: comma).
Returns: CSV text content.
Inputs: excel_base64 (required), sheet_index, delimiter. Returns: result
excel_to_jsonA
Extract data from an Excel sheet and return structured JSON.
Returns headers (if present) and typed data rows.
Args: excel_base64: Base64-encoded Excel file. sheet_index: Sheet index to extract (0-based, default: 0). has_header: Whether the first row is a header (default: true).
Returns: JSON with sheetName, totalRows, headers, and data array.
Inputs: excel_base64 (required), sheet_index, has_header. Returns: result
export_receipts_csvA
Export extracted receipt data as DATEV-compatible CSV.
Args: receipts: List of receipt extraction results (from extract_receipt).
Returns: Base64-encoded CSV file.
Inputs: receipts (required). Returns: result
export_receipts_xlsxA
Export extracted receipt data as Excel workbook (XLSX).
Args: receipts: List of receipt extraction results (from extract_receipt).
Returns: Base64-encoded XLSX file.
Inputs: receipts (required). Returns: result
extract_receiptA
Extract structured data from a receipt or invoice image using AI.
Returns vendor, date, totals (gross/net/VAT per rate), line items, currency, payment method, SKR03/04 account, category, and confidence score.
Args: image_base64: Base64-encoded receipt image (JPEG, PNG) or PDF. filename: Original filename (helps with format detection). content_type: MIME type (default: image/jpeg). Use application/pdf for PDF receipts.
Returns: JSON with extracted receipt data.
Inputs: image_base64 (required), filename, content_type. Returns: result
extract_receipt_asyncA
Submit a receipt for async AI extraction. Returns a job ID for polling.
Use get_receipt_job to check status and get_receipt_job_result for the result.
Args: image_base64: Base64-encoded receipt image or PDF. filename: Original filename. content_type: MIME type (default: image/jpeg). callback_url: Optional webhook URL to receive the result.
Returns: JSON with jobId and status.
Inputs: image_base64 (required), filename, content_type, callback_url. Returns: result
extract_text_from_pdfA
Extract all text content from a PDF.
Args: pdf_base64: Base64-encoded PDF file.
Returns: Extracted text content.
Inputs: pdf_base64 (required). Returns: result
extract_zugferdB
Extract structured invoice data from a ZUGFeRD/Factur-X PDF.
Args: pdf_base64: Base64-encoded ZUGFeRD PDF.
Returns: JSON with extracted invoice data.
Inputs: pdf_base64 (required). Returns: result
fill_excel_templateA
Fill an Excel template with data at named cells and ranges.
Args: template_base64: Base64-encoded Excel template file. values: Cell values to set (key = cell ref like 'Sheet1!A1' or named range, value = content). tables: Table data to insert at named ranges (key = range name, value = row arrays). recalculate: Recalculate formulas after filling (default: true).
Returns: Base64-encoded filled XLSX file.
Inputs: template_base64 (required), values, tables, recalculate. Returns: result
fill_pdf_formA
Fill form fields in a PDF.
Args: pdf_base64: Base64-encoded PDF file with form fields. fields: Field name-value pairs to fill. flatten: Whether to flatten the form (make fields non-editable).
Returns: Base64-encoded filled PDF.
Inputs: pdf_base64 (required), fields (required), flatten. Returns: result
generate_documentB
Generate a document with optional template, fields, and watermark.
Args: html_content: HTML content for the document body. markdown_content: Markdown content (alternative to HTML). template_name: Name of a pre-uploaded template (ODT/DOCX). fields: Template field replacements (key-value pairs). watermark: Diagonal watermark text overlay. output_format: Output format – PDF, DOCX, or ODT (default: PDF).
Returns: Base64-encoded document bytes.
Inputs: html_content, markdown_content, template_name, fields, +2 more. Returns: result
generate_excelA
Generate an Excel workbook (.xlsx) from a structured JSON definition.
The request dict should contain: sheets: List of sheet definitions, each with: name: Sheet tab name columns: Column definitions (header, width, format, align) rows: Data rows (list of dicts with 'values' or 'cells') formulas: List of formula definitions (cell, formula, label) headerFooter: Print header/footer configuration printArea: Print area in A1 notation freezePane: {row, col} for frozen panes headerStyle/dataStyle: Cell styling definitions autoSizeColumns, autoFilter, pageOrientation, fitToPage properties: Document properties (title, author, subject) password: Workbook protection password
Args: request: Structured JSON definition of the Excel workbook.
Returns: Base64-encoded XLSX file.
Inputs: request (required). Returns: result
generate_pdf_from_htmlC
Generate a PDF from HTML content.
Args: html: HTML content to convert to PDF.
Returns: Base64-encoded PDF bytes.
Inputs: html (required). Returns: result
generate_pdf_from_markdownA
Generate a PDF from Markdown content.
Args: markdown: Markdown content to convert to PDF.
Returns: Base64-encoded PDF bytes.
Inputs: markdown (required). Returns: result
get_pdf_metadataA
Get metadata (title, author, page count, etc.) from a PDF.
Args: pdf_base64: Base64-encoded PDF file.
Returns: JSON string with metadata fields.
Inputs: pdf_base64 (required). Returns: result
get_pdf_page_countB
Get the total number of pages in a PDF.
Args: pdf_base64: Base64-encoded PDF file.
Returns: Page count.
Inputs: pdf_base64 (required). Returns: result
get_receipt_jobA
Check the status of an async receipt extraction job.
Args: job_id: The job ID returned by extract_receipt_async.
Returns: JSON with job status (PENDING, PROCESSING, COMPLETED, FAILED).
Inputs: job_id (required). Returns: result
get_receipt_job_resultA
Get the extraction result of a completed async receipt job.
Args: job_id: The job ID returned by extract_receipt_async.
Returns: JSON with extracted receipt data.
Inputs: job_id (required). Returns: result
inspect_excelB
Inspect an Excel workbook and return metadata.
Returns sheet names, row/column counts, and named ranges.
Args: excel_base64: Base64-encoded Excel file.
Returns: JSON with sheetCount, sheets array, and namedRanges.
Inputs: excel_base64 (required). Returns: result
inspect_pdf_formA
Inspect form fields in a PDF and return their names, types, and current values.
Args: pdf_base64: Base64-encoded PDF file with form fields.
Returns: JSON array of form field definitions.
Inputs: pdf_base64 (required). Returns: result
list_certificatesB
List all uploaded signing certificates.
Returns: JSON array of certificate info objects.
Inputs: no named parameters. Returns: result
list_fontsA
List all uploaded custom fonts.
Returns: JSON array of font info objects.
Inputs: no named parameters. Returns: result
list_receipt_jobsB
List all async receipt extraction jobs.
Returns: JSON array of job info objects.
Inputs: no named parameters. Returns: result
list_templatesB
List all uploaded document templates.
Returns: JSON array of template info objects.
Inputs: no named parameters. Returns: result
merge_pdfsA
Merge multiple PDFs into a single PDF.
Args: pdfs_base64: List of base64-encoded PDF files to merge.
Returns: Base64-encoded merged PDF.
Inputs: pdfs_base64 (required). Returns: result
parse_xrechnungA
Parse XRechnung XML into structured invoice data.
Args: xml: XRechnung XML content.
Returns: JSON with parsed invoice data.
Inputs: xml (required). Returns: result
preview_pdf_pageA
Render a PDF page as a PNG image preview.
Args: pdf_base64: Base64-encoded PDF file. page: Page number (1-based, default: 1). dpi: Resolution in DPI (default: 150).
Returns: Base64-encoded PNG image.
Inputs: pdf_base64 (required), page, dpi. Returns: result
receipt_to_documentA
Extract receipt data and generate an expense report document.
Combines AI extraction with document generation in a single step.
Args: image_base64: Base64-encoded receipt image or PDF. filename: Original filename. content_type: MIME type (default: image/jpeg). output_format: Output format – PDF, DOCX, or ODT (default: PDF). title: Document title (default: "Spesenbeleg"). template_name: Optional template for the expense report layout.
Returns: JSON with extracted data and base64-encoded document.
Inputs: image_base64 (required), filename, content_type, output_format, +2 more. Returns: result
rotate_pdfA
Rotate pages in a PDF.
Args: pdf_base64: Base64-encoded PDF file. angle: Rotation angle in degrees (90, 180, 270). pages: Page range to rotate (e.g. "1-3", "all"). Default: all pages.
Returns: Base64-encoded rotated PDF.
Inputs: pdf_base64 (required), angle (required), pages. Returns: result
sign_pdfB
Digitally sign a PDF with a PKCS#12 certificate.
Args: pdf_base64: Base64-encoded PDF to sign. certificate_name: Name of a pre-uploaded certificate. certificate_password: Password for the certificate. reason: Reason for signing. location: Location of signing.
Returns: Base64-encoded signed PDF.
Inputs: pdf_base64 (required), certificate_name (required), certificate_password (required), reason, +1 more. Returns: result
upload_templateA
Upload a document template (ODT or DOCX).
Args: template_base64: Base64-encoded template file. filename: Filename for the template (e.g. "invoice.odt").
Returns: JSON with upload result.
Inputs: template_base64 (required), filename (required). Returns: result
validate_xrechnungA
Validate XRechnung XML for compliance.
Args: xml: XRechnung XML content.
Returns: JSON validation result (valid, format, errors, warnings).
Inputs: xml (required). Returns: result
validate_zugferdA
Validate a ZUGFeRD/Factur-X PDF for compliance.
Args: pdf_base64: Base64-encoded PDF with embedded ZUGFeRD data.
Returns: JSON validation result (valid, profile, errors, warnings).
Inputs: pdf_base64 (required). Returns: result
verify_pdf_signaturesB
Verify digital signatures in a PDF.
Args: pdf_base64: Base64-encoded signed PDF.
Returns: JSON with verification results (signed, signatureCount, allValid, signatures).
Inputs: pdf_base64 (required). Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 40 tools
Every tool has a clearly distinct purpose, covering document generation, PDF manipulation, Excel, receipts, e-invoicing, and template management. Overlaps like generate_document vs compose_document are resolved by different input types (single vs multi-part).
All tools follow a consistent verb_noun pattern with underscores, e.g., create_invoice, extract_receipt, merge_pdfs. No mixing of conventions like camelCase.
At 40 tools, the server is significantly over-scoped for a typical document generation service. The breadth of features (PDF, Excel, receipts, e-invoicing) is excessive, making it harder to navigate.
The tool set covers the full lifecycle for document generation, PDF operations, Excel handling, receipt extraction, and e-invoicing (ZUGFeRD/XRechnung). Includes CRUD for templates, async jobs, and validation.
Maintenance
Related MCP Connectors
PDFs, images and e-invoices (XRechnung, ZUGFeRD) from templates and JSON, rendered in the EU.
Convert, merge, compress and check PDFs, images, Office files and e-invoices (XRechnung, ZUGFeRD).
German e-invoicing: generate XRechnung 3.0, pre-check and extract (incl. ZUGFeRD PDF). Paid.
XRechnung and ZUGFeRD e-invoicing (EN 16931): create, validate, check Leitweg-IDs, German VAT.
Related MCP Servers
AlicenseNot gradedqualityCmaintenanceEnables creating professional documents (invoices, contracts, certificates, proposals, reports) via the DocuQueue API, with tools for template management, filling, previewing, and PDF generation.28 npm1MIT
rechnungsapi-mcpofficial
AlicenseAqualityBmaintenanceEnables AI assistants and agents to create, validate, read and convert German and EU e-invoices in ZUGFeRD, XRechnung and Factur-X formats, including extracting structured data from scanned or PDF invoices and embedding XML into PDF/A-3 files. It can run as a hosted remote server or locally, with each request authenticated by the caller's own API token.1024 npmMIT- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to read, validate against the official KoSIT rules (no Java), create, and display XRechnung and ZUGFeRD e-invoices, and export incoming invoices as DATEV posting batches. Running it with a cloud language model sends invoice contents to that provider's service.MIT

pdfvera MCP serverofficial
AlicenseAqualityBmaintenanceEnables AI assistants to convert German PDF invoices into XRechnung 3.0 and ZUGFeRD e-invoices with KoSIT validation and missing-field prompts, and Brazilian bank statement PDFs into OFX, CSV, and Excel after balance validation.3740 npmMIT