About invoiceproof
invoiceproof is a local bridge to the eInvoiceChecker hosted MCP service for processing European electronic invoices. It forwards tools for format detection, parsing, validation, deterministic fixes, and HTML or PDF rendering across supported standards.
Key features
- 01List supported invoice formats and rule sets
- 02Detect invoice format and profile
- 03Parse invoices into normalized JSON
- 04Validate invoices and repair deterministic errors
- 05Render invoices as HTML or PDF
Use cases
- 01Identify an incoming invoice’s format before validation
- 02Extract invoice fields for downstream processing
- 03Check invoice compliance and correct deterministic issues
- 04Generate a readable HTML or PDF invoice
MCP support
Exposes eInvoiceChecker’s invoice-processing tools through a local stdio MCP server, forwarding calls and results to the upstream service.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
8Saved tool list · updated 1w ago
This tool takes no parameters.
Details
From the project READMEinvoiceproof
invoiceproof is a thin local bridge to the eInvoiceChecker hosted MCP server, a service for working with European electronic invoices. It can tell you which formats are supported, identify an invoice you hand it, extract its contents as normalised JSON, check it against the official rule set for its format, repair the deterministic errors it finds, and render the document as HTML or PDF.
The bridge holds no invoice logic and validates nothing itself. It serves the upstream server's tool list locally — exactly as captured from the live endpoint — and forwards each tool call upstream, returning the result untouched. Coverage spans XRechnung, ZUGFeRD and Factur-X, Peppol BIS, KSeF, and the other formats and national rule sets the upstream supports.
Tools
No tool declares a required field; the invoice goes
in as content (XML text) or content_base64 (a
ZUGFeRD / Factur-X PDF or other file), both optional
in the captured schemas.
list_supported_formats— which formats, profiles and rule sets the service supports, and what each tool costs. No arguments.detect_invoice— identify the format, syntax, profile/customisation and country standard of an invoice file. The cheap first step before validating.parse_invoice— extract the invoice as normalised JSON: number, dates, seller, buyer, lines, VAT breakdown, totals and payment details.fix_invoice— deterministically repair validation errors of an EN 16931 invoice: empty elements, decimals, date and code formats, specification ids; withtotals: trueit also recalculates totals and the VAT breakdown from the lines. Data missing from the file can be supplied undervalues; anything still missing is listed inneeds_input.dry_runreports changes without applying them. Counts as 2 requests upstream.validate_invoice— run the official rule set for the detected format and return every failed rule with id, severity, message and XPath.kosit: falseskips the full KoSIT schematron for a faster, schema-and-core-only check.render_invoice_html— a human-readable invoice document in HTML (A4 print layout), with label language selectable vialang(en,de,fr,pl,nl,hr,es).render_invoice_pdf— the same document as a PDF, returned as a base64 resource, with the samelangchoices. Counts as 2 requests upstream.my_usage— the plan, monthly quota, usage this month and subscription state of the API key used for the connection. Free. No arguments.
Configuration
EINVOICECHECKER_API_KEY(optional) — when set, every upstream request carriesAuthorization: Bearer <key>. When it is not set, no auth header is sent and the upstream's anonymous rate limit applies. The key is never required by this client.EINVOICECHECKER_MCP_URL(optional) — override the upstream endpoint. Defaults tohttps://einvoicechecker.com/v1/mcp. (The override hook is this rebuild's own labelled reconstruction.)
Forwarding mechanics
On the first tool call the client opens a streamable
HTTP session with the upstream: a JSON-RPC
initialize request (protocolVersion 2025-03-26,
clientInfo name invoiceproof), then a
notifications/initialized notification, then the
tools/call request carrying the caller's arguments
unchanged. When the upstream answers initialize
with an Mcp-Session-Id header, that value rides on
every later request of the session, which is reused
for the life of the process. Replies may arrive as
plain JSON or as server-sent events; both are read.
Successful upstream results pass through verbatim,
including any isError flag the upstream sets.
Failures — a JSON-RPC error, a non-2xx status, an
unreadable response, or a request that cannot be
completed — come back as an isError result whose
text starts with invoiceproof error: (this
rendering is this rebuild's own labelled
reconstruction).
CLI
Each tool is also a subcommand taking one JSON argument string (omit it for tools that take none):
npx @reneturcios/invoiceproof list_supported_formats
npx @reneturcios/invoiceproof validate_invoice '{"content":"<Invoice>...</Invoice>"}'
npx @reneturcios/invoiceproof render_invoice_html '{"content":"<Invoice>...</Invoice>","lang":"de"}'
npx @reneturcios/invoiceproof --version
Exit codes: 0 on success, 1 on an isError
result, 2 on a usage error.
MCP server
npx -y @reneturcios/invoiceproof mcp
MCP client config:
{
"mcpServers": {
"invoiceproof": {
"command": "npx",
"args": ["-y", "@reneturcios/invoiceproof", "mcp"]
}
}
}
To run the MCP server with an API key, add
EINVOICECHECKER_API_KEY to the client's env
block for this server.
Docker
docker run -i ghcr.io/cloudcorprecords/invoiceproof
docker run --rm ghcr.io/cloudcorprecords/invoiceproof --version
The container's default command is mcp, so the first
form is a stdio MCP server; any other arguments reach
the CLI normally.
Development
npm install
npm run build
npm test
npm run dev
The test suite runs entirely against a stubbed JSON-RPC upstream — no live network calls.
License
MIT — see LICENSE.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.