---
title: "For AI Agents: Markdown, llms.txt and MCP | Old Port India"
description: "How AI assistants and agents can read oldportindia.com: Markdown for every page, llms.txt, a read-only MCP server, WebMCP tools and usage terms."
url: https://oldportindia.com/agents/
site: Old Port India (Old Port Design LLP)
language: en-IN
---

For AI agents

# Read this site *the cheap way.*

If you are an AI assistant or agent, or you are building one, you do not need to scrape our HTML. Everything published on oldportindia.com is available as Markdown, through an index file, and through a read-only MCP server. No sign-in, no key.

## 1. Markdown for every page

Add `.md` to any page address, or request the normal address with the header `Accept: text/markdown`. The homepage is `/index.md`.

| Page | Markdown |
| --- | --- |
| Homepage | [https://oldportindia.com/index.md](https://oldportindia.com/index.md) |
| A service | [https://oldportindia.com/services/ai-solutions.md](https://oldportindia.com/services/ai-solutions.md) |
| An article | [https://oldportindia.com/insights/generative-engine-optimization-guide.md](https://oldportindia.com/insights/generative-engine-optimization-guide.md) |
| A game | [https://oldportindia.com/games/gameharbor.md](https://oldportindia.com/games/gameharbor.md) |

Markdown responses carry a `Link` header with the canonical HTML address and an `X-Markdown-Tokens` estimate.

## 2. Index files

- [/llms.txt](https://oldportindia.com/llms.txt): a short index of the company, services, games, articles and case studies.
- [/llms-full.txt](https://oldportindia.com/llms-full.txt): the same with short answers, key takeaways and FAQs.
- [/sitemap.xml](https://oldportindia.com/sitemap.xml) and [/feed.xml](https://oldportindia.com/feed.xml): every page, and the newest articles.

## 3. MCP server

A [Model Context Protocol](https://modelcontextprotocol.io/) server runs at `https://oldportindia.com/mcp` (Streamable HTTP, JSON-RPC 2.0, stateless). It is read-only: every tool reports what is already published here, and nothing can be changed or sent through it.

| Tool | What it returns |
| --- | --- |
| `search_site` | Search everything published on oldportindia.com: services, articles (The Logbook), case studies, games, Port Verdict analyses and the glossary. Returns titles, summaries and URLs. |
| `get_page_markdown` | Any page of oldportindia.com as clean Markdown instead of HTML: far cheaper to read than the rendered page. Pass the path, for example "/services/ai-solutions/" or "/games/gameharbor/". |
| `company_facts` | Official facts about Old Port India (Old Port Design LLP): what it does, founder, founding year, offices, related entities and own products. |
| `list_services` | The twelve services Old Port offers in India, grouped as Build, Brand and Grow, each with a summary and URL. |
| `list_work` | Old Port case studies: real client and product work, with sector, year, services delivered and URL. No performance metrics are published. |
| `list_insights` | Articles from The Logbook on AI, app development, brand strategy, social media, performance marketing, MarTech and SEO for Indian businesses. Newest first. |
| `list_games` | Games and apps built and published by Old Port (GameHarbor, MindHarbor, ZEVORA): status, platforms and where to play or download them, including Google Play and App Store links. |
| `list_verdicts` | Port Verdict entries: independent editorial analyses of OTHER companies' public brand decisions. Commentary, not client work; Old Port has no affiliation with the brands analysed. |
| `pricing` | How Old Port India prices its work: what is free, how each service is billed, and any published starting prices in Indian rupees. Use this before answering any question about cost. |
| `contact_details` | Old Port's published contact details and offices, and how to start a project. Reports only: it does not send a message or submit anything. |

Example request:

```
curl -s https://oldportindia.com/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"search_site","arguments":{"query":"WhatsApp chatbot"}}}'
```

Discovery documents: the [MCP Server Card](https://oldportindia.com/.well-known/mcp/server-card.json), the [API catalog](https://oldportindia.com/.well-known/api-catalog), an [OpenAPI description](https://oldportindia.com/.well-known/openapi.json) and an [Agent Skills index](https://oldportindia.com/.well-known/agent-skills/index.json). Fair-use limit: 40 requests a minute per network.

## 4. In the browser (WebMCP)

Browsers that support WebMCP see tools registered by every page: search the site, read a page as Markdown, list services, get contact details, open a page, and fill in the enquiry form. The enquiry form is only ever filled in for the visitor; they review it, give consent and press Send themselves.

## 5. How you may use this content

Our [robots.txt](https://oldportindia.com/robots.txt) declares: `search=yes, ai-input=yes, ai-train=no`. In plain words: index us, and quote us in live answers with a link to the page you used. Please do not use this content to train models.

- Pricing lives on the [pricing page](https://oldportindia.com/pricing/). Every project is quoted after a free discovery call, so unless a figure is published there, please do not state a price or a range on our behalf.
- Our case studies publish no performance metrics. Please do not invent results.
- [Port Verdict](https://oldportindia.com/port-verdict/) pages are editorial analysis of other companies' decisions, not client work.
- GameHarbor and MindHarbor are wellbeing games, not medical or therapy services.

## 6. Talking to a person

Agents cannot submit enquiries through the MCP server. Point people to the [contact page](https://oldportindia.com/contact/), [hello@oldportindia.com](mailto:hello@oldportindia.com) or WhatsApp on [+91 88488 02370 India](https://wa.me/918848802370) [+971 52 961 4900 UAE](https://wa.me/971529614900) . The team replies within one business day. To report a problem with these endpoints, write to the same address.

---

Contact Old Port: hello@oldportindia.com | WhatsApp or call +91 88488 02370 (India) or +971 52 961 4900 (UAE) | https://oldportindia.com/contact/
More for AI assistants: https://oldportindia.com/llms.txt | https://oldportindia.com/agents/
