Skip to main content
Glama
Scholar Relay

Scholar Relay

by adhithyan05

Scholar Relay

An AI-powered research assistant built using NitroStack MCP and LangGraph that simplifies academic research by helping students discover research papers, generate summaries, build bibliographies, and continue their research seamlessly across multiple devices.


🚀 Features

  • 🔍 Discover relevant research papers based on a topic

  • 📄 AI-generated paper summaries

  • 📚 Automatic bibliography generation

  • 📱 Cross-device session continuation using QR codes

  • 📝 Citation verification and unsupported claim detection

  • 🎓 Supports multiple citation styles (APA, IEEE, MLA)

  • 🤖 Agentic AI workflow powered by LangGraph

  • ⚡ Built on NitroStack MCP with Tools, Resources, and Prompts


Related MCP server: ScholarMind

🏗 Tech Stack

  • NitroStack MCP

  • LangGraph

  • TypeScript

  • React

  • Tailwind CSS

  • Node.js

  • Zod

  • OpenAI API


🧠 MCP Components

Tools

  • discover_papers

  • summarize_paper

  • check_citations

  • generate_handoff_token

  • notify_readiness_report

Resources

  • bibliography://{token}

  • catalog://papers

  • policy://citation-style

Prompts

  • Research Assistant

  • Citation Style Formatter

  • Report Generator


📋 Project Workflow

  1. Student enters a research topic.

  2. AI discovers relevant papers.

  3. Papers are summarized.

  4. A bibliography is automatically created.

  5. The bibliography is stored as an MCP Resource.

  6. A QR code is generated for session handoff.

  7. Research continues seamlessly on another device.

  8. Draft paragraphs are checked for unsupported claims.

  9. Missing citations are suggested.

  10. A final research readiness report is generated.


⚙️ Installation

Clone the repository

git clone https://github.com/adhithyan05/scholar-relay.git

Go to the project folder

cd scholar-relay

Install dependencies

npm install

Run the development server

npm run dev

🎯 Use Cases

  • University Libraries

  • Research Institutions

  • Students

  • Faculty Members

  • Literature Reviews

  • Thesis and Dissertation Writing


🌍 Future Enhancements

  • Integration with arXiv and Semantic Scholar

  • Real-time collaboration

  • AI-powered plagiarism prevention

  • Personalized research recommendations

  • Cloud synchronization


👥 Team

  • Adhithyan S

  • Goutham krishna VG

  • Devadath Krishna

  • Abhinav P Madhu


📜 License

This project was developed as part of the NitroStack MCP Hackathon 2026 under the Education & Research track.


🙏 Acknowledgements

  • NitroStack

  • LangGraph

  • OpenAI

  • Amrita Vishwa Vidyapeetham

Available Tools

11 tools
add-noteB

Add a textual note to your persistent research session.

Inputs: note (required).

check-citationsA

Analyze a draft paragraph and flag claims that are unsupported, overly absolute, or contradicted by the paper library. Returns each flagged claim with a reason and suggested sources.

Inputs: paragraph (required).

clear-sessionA

Clear all data in the current research session, resetting it completely.

Inputs: no named parameters.

compare-papersA

Compare 2-5 research papers side-by-side. Returns a structured JSON comparison matrix.

Inputs: paperIds (required).

discover-papersA

Search the arXiv library for a research topic and return matching paper cards. Renders a browsable grid.

Inputs: topic (required), limit.

find-research-gapsA

Analyze the latest arXiv papers on a topic to identify under-explored areas, open problems, and novel project ideas.

Inputs: topic (required).

generate-literature-reviewA

Generate a comprehensive AI literature review on a given topic by searching arXiv, summarizing top papers, and comparing them.

Inputs: topic (required).

get-current-sessionA

Retrieve the entire persistent research session, including the current topic, summarized papers, literature reviews, gaps, notes, and saved papers. This opens the Research Workspace widget.

Inputs: no named parameters.

readiness-reportA

Given a list of arXiv paper ids that make up a bibliography, produce a citation-readiness summary: which entries resolve, coverage by topic, and any missing ids.

Inputs: paperIds (required).

save-paperB

Save a paper to your persistent research session for later reading or reference.

Inputs: id (required), title (required), arxivUrl (required).

summarize-paperA

Summarize a single paper by its arXiv id into a structured summary. Use the id returned by discover-papers.

Inputs: paperId (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.9/5.0

    Scored across 11 tools

    Disambiguation4/5

    Most tools have clear, distinct purposes, but generate-literature-review appears to encompass discover-papers, summarize-paper, and compare-papers, creating potential overlap. check-citations and readiness-report both relate to citation readiness, though they operate on different inputs (a paragraph vs. a bibliography).

    Naming Consistency4/5

    The vast majority of tools follow a verb_noun pattern (e.g., discover-papers, summarize-paper, save-paper). However, readiness-report deviates by using a noun_noun structure, breaking the otherwise consistent convention.

    Tool Count5/5

    With 11 tools, the server is well-scoped for a research assistant domain. Each tool serves a distinct function in the research workflow, from discovery to citation analysis, without feeling bloated or sparse.

    Completeness4/5

    The surface covers the core research lifecycle: search, summarize, compare, generate reviews, find gaps, and manage session state. However, there is no way to remove or update a single saved paper or note, only clear the entire session, which is a notable gap for persistent session management.

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      A LangGraph-powered research agent that performs iterative web searches using Google Search and Gemini models to generate structured reports with citations. It integrates with MCP-compatible clients like Claude and Cursor to enable sophisticated, multi-step AI research workflows.
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      Enables searching arXiv and top AI conferences, finding related papers, generating research insights, and managing a personal library via MCP tools.
      5
      50 npm
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      A production-quality multi-agent research system with an arXiv MCP server and LangGraph-based research agents, providing search, details, and cached summaries via SQLite.
      MIT