NTU Library MCP
AI research sidekick for NTU students

Find the right source.
Skip 200 abstracts.

A search still hands you hundreds of hits — and the real chore is opening each abstract to see if it actually fits. This plugs NTU's library into your AI so it reads the abstracts for you: which sources fit, what they argue, and which copy you can download. Essay, a single claim, or just a reference with a real link.

Set it up in 2 min Explore use cases
live · knowledge graph drag to explore — every node is a source
0
abstracts you'd otherwise skim by hand
0
relevant picks it hands back — abstracts already read
0
free & open source, forever
use it for

One tool, a dozen reasons

The CC0001 op-ed is just where it started. Any time you need a source, a reference, or a link — this is the move. See the full Playbook →

“Give me a credible source I can cite for this claim — with a link I can open.”

back up a statement

“Find info on urban heat islands and give me the reference.”

look something up

“Any recent papers on this I can actually download tonight?”

get the full text

“What's the most-cited work on this topic — and can NTU get it?”

find the key paper

“Build me a short reading list for my essay on ___.”

start a project

“Is this book in the NTU library, and where do I find it?”

locate a title
Explore all use cases in the Playbook
what it does

Four moves, so you stop doomscrolling results

01 · search

Ask in plain English

“Find peer-reviewed sources on hawker culture.” Your AI hits NTU OneSearch directly — no login wall, no facet checkboxes.

02 · read

Abstracts, pre-read

Every hit returns its abstract, author, year and DOI — so the AI tells you what a paper argues before you open a tab.

03 · rank

Find the influential work

A second tool taps Crossref's 150M-paper index to surface the most-cited research on your topic — so you cite what the field actually builds on.

04 · decide

Which copy to grab

The real time-saver: each source is tagged download free, online via NTU login, or print only. No more dead links.

what's included

Everything in the box

search_ntu_library

What NTU holds. Searches OneSearch (Primo) and returns title, author, year, abstract, DOI, a permalink, and an access tag per result.

querylimit?resource_type?

find_influential_research

What the field cites most. Taps Crossref's ~150M-paper open index to surface high-impact work on a topic, then you cross-check what NTU can give you.

querylimit?from_year?sort?

get_ntu_record

Deep-read one item. Full abstract and access detail for a single record by its Primo id.

record_id

Smart citation ranking (not naïve)

Crossref's own sort=citations returns the most-cited paper on Earth regardless of topic. So this fetches a relevance-matched pool first, then ranks that by citations — you get work that's influential and on-topic. A hawker-culture query surfaces “Singapore Hawker Centers: Origins, Identity, Authenticity”, not “Global Burden of Disease”.

where it started

It began with one op-ed. Now it's for any time you need a source.

By an NTU student  ·  Started in CC0001, Inquiry & Communication

The spark was the CC0001 op-ed — the note my tutor kept leaving: “support this with a credible source.” So I'd open OneSearch, type a topic, and get two million results back. Helpful.

But that problem isn't an op-ed problem. It's every assignment, every claim, every “wait, what do I even cite for this?”

You rarely need two million results. You need the few that are actually relevant, an abstract you can skim in ten seconds, and a copy you can open tonight — whether you're writing an essay, backing up a single sentence, or just chasing down where a fact comes from.

AI is great at that triage. It just couldn't see NTU's catalogue — until now. This is a small, free bridge: it hands your AI the library's public search and the field's citation index, and asks it to do the boring part. Find the relevant few. Read what they argue. Hand you the link.

It's open source, it stores nothing about you, and it's for anyone at NTU who's ever needed a source and a link — fast. That's the whole pitch.

one-click setup · ~2 min

Point your AI at the library

No files to edit. In Claude Desktop, claude.ai, or the mobile app:

  1. Open Settings → Connectors.
  2. Click Add custom connector.
  3. Paste the server URL below and hit Add. That's it.
server URL — paste this
https://ntu-library.openskillshub.org/mcp
Official guide Watch a walkthrough

Works on Free, Pro, Max, Team & Enterprise (Free = 1 custom connector).

In ChatGPT (Plus, Pro, Business, Enterprise or Edu — not Free):

  1. Open Settings → Connectors (on Business/Enterprise, an admin first enables Developer mode under Connectors).
  2. Click Create / Add custom connector.
  3. Paste the server URL below, name it “NTU Library”, and save.
server URL — paste this
https://ntu-library.openskillshub.org/mcp
Connectors in ChatGPT Watch a walkthrough Developer mode

Custom connectors aren't on the Free plan. Business/Enterprise need an admin to turn on Developer mode.

Use the Gemini CLI — one command adds it (no manual file edits):

terminal
gemini mcp add -t http ntu-library https://ntu-library.openskillshub.org/mcp
Gemini CLI — MCP guide Watch a walkthrough Gemini Enterprise

On Gemini Enterprise, an admin connects a custom MCP data store (see the Enterprise link). The consumer Gemini app doesn't take custom MCP servers yet.

One command in your terminal:

terminal
claude mcp add --transport http ntu-library https://ntu-library.openskillshub.org/mcp
Claude Code — MCP docs Watch a walkthrough

Add to ~/.cursor/mcp.json (or any MCP client that speaks Streamable HTTP):

~/.cursor/mcp.json
{
  "mcpServers": {
    "ntu-library": {
      "url": "https://ntu-library.openskillshub.org/mcp"
    }
  }
}
Cursor — MCP docs Watch a walkthrough
Added it? Here's what to do next.The Playbook has copy-paste prompts for every kind of research job.
Open the Playbook
see it in action

A 60-second demo

Watch it run inside Claude Code — or read the exchange below.

You

I'm arguing that Singapore's hawker culture is worth preserving. Find me credible sources I can actually read and cite.

Claude + NTU Library

Found 3 strong picks in NTU OneSearch (out of 18,400):

“Hawker centres as social infrastructure” (2021) — J. Tan · Open access — download directly, no login.
“Intangible heritage & urban food” (2019) ·
“Preserving Singapore's food streets” (2016) · Print only — on the shelf at Lee Wee Nam Library.

Want me to pull the most-cited paper on this too, so your argument leans on the field's key work?

the fine print, honestly

What it will & won't do

Reads the public catalogue

Only NTU's public Primo search plus open scholarly metadata (Crossref). The same results you'd get in a browser. Nothing behind a paywall.

Won't pirate anything

No full-text scraping, no EZproxy wrapping, no re-hosting PDFs. It points you to the legit copy; you grab it the normal way.

Stores nothing about you

Usage counts are anonymous and hashed — no IPs, no names. Just enough to know people use it.

Free & open source

Runs on Cloudflare's edge. For NTU students, by an NTU student. Fork it, read it, improve it.

Copied to clipboard