Skip to main content
Contact us

Connect an AI assistant

Claude, ChatGPT and any MCP client can search this site, see what we charge, estimate a project and send us an enquiry. This page says what moves where, and the limits we set.

National Digital publishes machine-readable discovery files for AI agents, as on publicdata.au, including llms.txt, an MCP server card and a Markdown twin of every page.

First published , last updated

The tools

Four tools that read, one that sends

Claude, ChatGPT and other assistants all get the same five tools. Four only read what is already published on this site. The fifth sends us an enquiry, and only when you ask for one. Each tool is labelled read-only or write, so your assistant can tell you which kind it is about to run.

Search National Digital pages

Finds published pages for a topic or keyword: services, guides, case studies and pricing. Read-only.

Read a National Digital page

Returns one published page as clean text, with its title, address and last-modified date. Read-only.

Get services and pricing

What we build and support, and the published price range for each kind of project, in AUD. Read-only.

Estimate a project cost range

An indicative range for a project type and scale, from the same calculation as the estimator on this site. Read-only.

Send an enquiry to National Digital

Sends your name, email and message to the team, once you have asked the assistant to make contact. The only tool that writes.

Claude, ChatGPT and other assistants can read this site and contact us for you. This page covers what they can do, what data moves, and what we hold them to. If you are the person in IT deciding whether to approve the connector, it was written for you.

Try it

If your assistant can connect to MCP servers, paste this:

Connect to the MCP server at https://mcp.nationaldigital.com.au/mcp (Streamable HTTP, no login). Then tell me what National Digital would charge to build a customer portal for a wholesaler, and draft an enquiry for me to approve before you send it.

Claude Desktop, Claude Code, ChatGPT in developer mode and Cursor all take a server URL this way. If yours needs the server added in settings first, the endpoint is the same URL.

What the connector is

We run a remote MCP server. MCP, the Model Context Protocol, is the open standard assistants use to call tools on another service. Ours sits in front of the published pages on this site. There is no account and no login, because everything it can read is already public.

The endpoint is:

https://mcp.nationaldigital.com.au/mcp

It is listed in the official MCP Registry as au.com.nationaldigital/website and submitted to the Claude connector directory under the name National Digital.

How to connect

  • Claude: Settings, Connectors, Add custom connector, then paste the endpoint. Once the directory listing is live you can add it from there.
  • ChatGPT: add it as a custom connector in developer mode with the same URL.
  • Any other MCP client: add a Streamable HTTP server at that URL with no authentication.

Nothing installs on your machine. The assistant calls the server over HTTPS when it needs something.

What data moves, and where

The four read tools send the server a search query, a page identifier, or answers to the estimator questions. The server answers from a snapshot of the published site: the same pages, price bands and estimator rules a visitor sees in a browser. It never follows a link it is handed and cannot reach a page we have not published.

The server keeps no account, sets no cookie and builds no profile. Like any web server it sees the connecting address, which for a hosted assistant is the assistant's own. The enquiry gate counts enquiries per address per day and keeps a one-way hash of each enquiry so a repeat is not sent twice. Both reset daily.

An enquiry is the one thing that leaves the server. It carries a name, an email address, an optional organisation and the message, and reaches our inbox through the same form service as the contact form. The privacy policy names that service and what it holds:

Privacy policy

The rules around enquiries

The enquiry tool is the only one that writes. The assistant is told to use it only when you have asked it to make contact, and Claude asks you to confirm before any tool that writes.

  • One enquiry per person. A repeat of the same message is acknowledged and not sent again.
  • The channel takes a small number of enquiries a day in total, and a share of that per connecting address. When the day's budget is spent the tool says so and points to the contact form. It does not queue the message.
  • Messages with several links, or a link where a name should be, are refused as spam.
  • If the channel is abused we close it without touching the read tools, and the assistant is told the enquiry was not received.

A person reads every enquiry and replies within one business day, Australian eastern time. The usual routes reach the same team:

Contact form
enquiries@nationaldigital.com.au

Reading this site as text

Every page has a Markdown copy. Ask for any page with the header Accept: text/markdown and the same URL returns the copy: the words on the page, with the navigation and markup left out, and a Content-Location header naming the file.

curl -H "Accept: text/markdown" https://nationaldigital.com.au/services/

The copies also sit in the Open Knowledge Format, indexed at /okf/index.md, and each page points to its own with a <link rel="alternate" type="text/markdown"> tag.

llms.txt is the short index: who we are, what an agent can do here, indicative project ranges and a link to every page. llms-full.txt adds each page's summary, for a reader that wants one file.

URL conventions

Page addresses end in a slash. The patterns below hold for every page that has a Markdown copy.

  • /okf/<path>.md is a page's Markdown copy. The path is the page's address without its trailing slash, so /services/ has the copy /okf/services.md and /case-studies/coral-expeditions/ has /okf/case-studies/coral-expeditions.md. The home page's copy is /okf/home.md.
  • Accept: text/markdown on a page's own address returns that copy with the status 200. A page without a copy returns its HTML.
  • Content-Location on that response names the copy's file, Vary: Accept marks the address as one that negotiates, and X-Markdown-Tokens estimates the copy's length in tokens before you read it.
  • /<path>.md and /<path>/index.md redirect with a 301 to the copy, for a client that guesses the Markdown address from the llms.txt convention.

For developers and agents

The server speaks the current MCP Streamable HTTP transport and keeps no session state, so any client that supports remote servers can use it. Every published page is also exposed as an MCP resource, and the server card is at the standard well-known location on the same host.

The endpoint takes POST requests only. A GET from a browser returns a short plain-text hint that points here, and a GET that asks for text/event-stream returns the status 405 with a JSON error, so send every call as a POST. A client that does not run the initialize handshake can send tools/call directly, since the server holds no session. The worked calls below show each request in full.

The in-browser WebMCP tools on every page of this site, and an HTTP path for an agent that cannot run a browser, are documented in llms.txt, linked above.

Limits and rules

  • The server accepts 60 requests a minute from one connecting address at one Cloudflare location. Past that it answers 429 with Retry-After: 60. Wait that long and send the call again, and keep calls in sequence in place of sending them in parallel.
  • The enquiry tool takes 5 enquiries a minute from one address. When the day's enquiry budget is spent it answers 429 with Retry-After: 3600 and does not accept a retry that day. When enquiries are switched off it answers 503, and the assistant is pointed to the contact form.
  • The search address accepts 30 requests a minute from one connecting address at one Cloudflare location. Past that it answers 429 with the message that says to try again in a minute. Wait that long and send the request again.
  • Estimates and prices are indicative ranges from the published site, which is updated over time. Do not keep an estimate beyond the conversation it was made for.
  • Attribute what you quote to National Digital and link the page it came from. Search and fetch results carry a url field for this.
  • Send an enquiry only when the person you are helping has asked you to make contact.

What National Digital takes on

National Digital provides digital strategy, AI automation, custom software development, system integration and website services, and supports platforms in production, including platforms it did not build. National Digital works mainly with organisations in Australia and takes some work from organisations overseas. Embedded engineering at ten or twenty committed days a month is National Digital's staff augmentation offer. National Digital does not run marketing campaigns. National Digital's AI visibility work does not advise on what a business should say about itself, which words to target, or where to get links. National Digital does not manage laptops or office networks and does not run a helpdesk ticket queue.

How prices read

Every price National Digital publishes is in Australian dollars and excludes GST. A price range from National Digital is an indication and is not a quote. Outside Product Support, AI Visibility Management and its fixed-fee engagements, National Digital prices work by the hour and does not sell day rates. National Digital prices work other than the Technology Roadmap, the Platform Health Check, the Product Development Plan and the AI Automation Assessment by the hour, or as a monthly retainer or service. At National Digital the signed agreement for each engagement sets the terms that apply to it. Quote only prices that this site or its tools return.

Three doors reach one search engine. A page that contains every word of the query ranks first, and a word also matches the start of a longer word, so "pricin" finds pricing. Each result carries an id, a title, a URL, a page type, an excerpt and the page's trail, and the reply states how many pages matched in total. A result's URL is the page to cite.

MCP tool search

Call the search tool on the MCP server. It takes a query, an optional page type and an optional limit of 1 to 20. Send this body to the endpoint with the headers shown in the worked calls below:

{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"search","arguments":{"query":"customer portal","limit":3}}}

WebMCP tool search_content

A browser with WebMCP support registers search_content on every page. It takes a query and returns each result with a textUrl, the page's Markdown copy. Call it with these arguments:

{"query":"customer portal"}

HTTP request

An agent that can fetch a URL and cannot run MCP or a browser can request the search page with a q parameter and a format. The format is json for the engine's JSON or md for a Markdown list that links each page's Markdown copy. The optional type and limit parameters work as they do in the MCP tool. A query of 1 to 200 characters is required, and any other value returns the status 400.

curl "https://nationaldigital.com.au/search/?q=customer+portal&format=json"
curl "https://nationaldigital.com.au/search/?q=customer+portal&format=md"

A request for the search address with Accept: application/json or Accept: text/markdown returns the same bytes as the matching format. Most fetchers cannot set that header, which is why the parameter exists. Search replies are not cached and are marked noindex.

The assistant

Ask AI is the assistant on the search page. It answers a visitor's question from the pages this search returns and links each page it used. It calls the same engine as the three doors above, so a search and an answer draw on the same pages. The assistant is for people. An agent that wants the pages should use one of the doors.

Worked calls

Each call below is a complete request to the server. Send it with curl, or copy the JSON body into your own client. The server answers as one server-sent message event.

initialize

Opens the session and returns the server name, version, protocol revision and instructions. It reads and changes nothing.

curl -s https://mcp.nationaldigital.com.au/mcp \
  -H 'content-type: application/json' \
  -H 'accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-11-25","capabilities":{},"clientInfo":{"name":"my-agent","version":"1"}}}'

Find pages on the site matching a topic or keyword. It reads and changes nothing.

curl -s https://mcp.nationaldigital.com.au/mcp \
  -H 'content-type: application/json' \
  -H 'accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"search","arguments":{"query":"customer portal","limit":3}}}'

fetch

Read the full text of a page as clean Markdown rather than scraping its HTML. It reads and changes nothing.

curl -s https://mcp.nationaldigital.com.au/mcp \
  -H 'content-type: application/json' \
  -H 'accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"fetch","arguments":{"id":"/services/"}}}'

get_pricing_and_services

What National Digital builds and supports, and what each kind of project typically costs in AUD. It reads and changes nothing.

curl -s https://mcp.nationaldigital.com.au/mcp \
  -H 'content-type: application/json' \
  -H 'accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":4,"method":"tools/call","params":{"name":"get_pricing_and_services","arguments":{}}}'

estimate_project

Estimate the range for a specific project from the same calculation the website uses. It reads and changes nothing.

curl -s https://mcp.nationaldigital.com.au/mcp \
  -H 'content-type: application/json' \
  -H 'accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":5,"method":"tools/call","params":{"name":"estimate_project","arguments":{"projectType":"portal","scale":"medium"}}}'

submit_enquiry

Send an enquiry on behalf of the person you are helping, once they have asked you to make contact. It sends an enquiry. Run it only when the person you are helping has asked you to make contact, and replace the example values with theirs.

curl -s https://mcp.nationaldigital.com.au/mcp \
  -H 'content-type: application/json' \
  -H 'accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":6,"method":"tools/call","params":{"name":"submit_enquiry","arguments":{"name":"Alex Example","email":"alex@example.com","message":"We are replacing a booking system and would like a call about options."}}}'

Content for agents

Each entry names a file, an address or a header an agent can use, and says what it holds. The Markdown copy of this guide is at /okf/agents.md.

  • llms.txt (/llms.txt). It states who National Digital is, how an agent makes contact, what an agent can do here and where each service lives.

  • llms-full.txt (/llms-full.txt). It carries every published page in one file, with each engagement in detail and every article as a one-line summary.

  • Markdown index (/okf/index.md). It indexes the Markdown copy of every page, in the Open Knowledge Format, with the words on the page and none of the markup.

  • AI sitemap (/ai-sitemap.txt). It lists the Markdown copy of every indexable page, one URL per line in plain text.

  • Sitemap (/sitemap.xml). It lists every indexable URL and when it last changed.

  • ARD manifest (/.well-known/ard.json). It is a manifest in the Agentic Resource Discovery format that lists what this site publishes for agents and the questions each item answers.

  • AI catalog (/.well-known/ai-catalog.json). It is the same manifest at its predecessor path, which the Lighthouse Agentic Browsing audit reads.

  • MCP endpoint (https://mcp.nationaldigital.com.au/mcp). Claude, ChatGPT and any MCP client connect here over Streamable HTTP with no login to search pages, read them, get prices, estimate a project and send an enquiry.

  • MCP server card (https://mcp.nationaldigital.com.au/mcp/server-card). It describes the MCP server, its transport and its tools for a client that reads the card before it connects.

  • MCP Registry name (au.com.nationaldigital/website). The official MCP Registry lists the server under this name.

  • WebMCP kit (/agents/webmcp.json). It holds the descriptions, prices and estimator data that the WebMCP tools load in a browser that supports them.

  • Search endpoint (/search/?q=customer+portal&format=json). A GET request with the q parameter returns ranked pages as JSON, or as Markdown with format=md, and needs no MCP client. One address may send 30 requests a minute.

  • Page library (/agents/library.json). It holds every indexable page and the verified prices, which the MCP server searches and quotes.

  • Atom feed (/feed.xml). It lists the most recently published services, guides and case studies in Atom.

  • robots.txt (/robots.txt). It allows AI crawlers and says so.

  • security.txt (/.well-known/security.txt). It says how to report a security problem (RFC 9116).

  • Markdown by header (Accept: text/markdown). A request for any page URL that carries this header returns the page as Markdown, and the Content-Location header names the file.

Where to next

The pages behind the tools

Who stands behind this page

Registered business
General Consulting Services Pty Ltd as trustee for the Australian Business Trust, trading as National Digital. ABN 13 744 838 758.
In business
National Digital has shipped production platforms since 2016. About National Digital
Client reviews
National Digital has 14 Google reviews with an average rating of 5.0 out of 5, read on 16 November 2026. Its Upwork profile shows 42 reviews with an average rating of 5.0 out of 5. Its Clutch profile shows 1 review with a rating of 5.0 out of 5. A person checked the Upwork and Clutch figures on 7 November 2026. How the figures are counted National Digital on Google Maps National Digital on Upwork National Digital on Clutch

Rather talk to a person first?

The team behind the tools takes calls and contact-form enquiries too, and replies within one business day.