# CrossRef MCP for AI Agents MCP

> CrossRef lets your AI client search over 150 million academic works in one place. You can find detailed metadata—like citation counts, authors, publication dates, and funding sources—for journal articles, books, and datasets instantly.

## Overview
- **Category:** the-unthinkable
- **Price:** Free
- **Tags:** academic-metadata, doi-lookup, scholarly-publishing, citations, research-tools

## Description

Need to track down scholarly details across massive databases? This MCP connects you directly to CrossRef, giving your AI client access to the world's largest repository of academic metadata. You don't need API keys or to navigate complex web forms; you just ask your agent a question in natural conversation.

Ask it for papers on climate modeling from 2018, and it finds them. Ask it to check which publishers are associated with biomedical journals, and it gives you the list. The result is structured data: titles, authors, DOIs, citation counts, and abstracts—all ready for your next task. If you're building a system that needs reliable, verifiable scholarly information, connecting via Vinkius allows any MCP-compatible client to access this power instantly.

## Tools

### get_funder
Retrieves specific details about a research funding organization, including their location and total works funded.

### get_journal
Gets comprehensive data for an academic journal using its ISSN, including publisher and work count.

### get_licenses
Provides a list of all known academic licenses used in scholarly works, along with their names and IDs.

### get_member
Retrieves details for a specific CrossRef member organization, including its prefixes and total work count.

### get_publisher
Gets information on an academic publisher, showing their name, work count, journal list, and location.

### get_types
Provides a full list of recognized publication types, useful for narrowing down search results.

### get_work
Pulls all metadata—title, authors, abstract, citation count, and journal details—for a single academic work given its DOI.

### get_works_by_doi_list
Gets metadata for multiple works at once by providing a comma-separated list of DOIs.

### search_funders
Searches across research funders to find names, IDs, and how many works they have supported.

### search_journals
Looks up academic journals by name to get their ISSNs, publisher details, and total work count.

### search_members
Searches the catalog of CrossRef member organizations by name or ID.

### search_publishers
Finds academic publishers by name to get their work count, journal list, and organization details.

### search_works
Performs comprehensive searches across the entire database using keywords, author names, or specific DOIs.

## Prompt Examples

**Prompt:** 
```
Find papers on large language models from 2023 and show me their citation counts.
```

**Response:** 
```
**Search Results: LLMs (2023)**

| Title | Authors | Journal | Year | Citations |
| :--- | :--- | :--- | :--- | :--- |
| The Transformer Architecture Revisited | Vaswani et al. | Nature Machine Intelligence | 2023 | 12,450+ |
| Context Window Limits | Smith & Jones | IEEE Trans. AI | 2023 | 8,900+ |

*Note: Citation counts are dynamic and retrieved live.*
```

**Prompt:** 
```
What kind of funding supported the work by DeepMind on protein folding?
```

**Response:** 
```
**Funding Details for AlphaFold:**

This research was primarily funded by:
*   **Organization Name:** National Institutes of Health (NIH)
*   **Country:** USA
*   **Total Works Funded:** 312
*   **Location:** Bethesda, MD

The NIH is a key contributor to this field.
```

**Prompt:** 
```
List the biggest academic publishers and how many journals they manage.
```

**Response:** 
```
**Top Academic Publishers by Scope:**

1.  **Elsevier:** 450+ Journals (Focus: Biomedicine)
2.  **Springer Nature:** 380+ Journals (Focus: STEM & Humanities)
3.  **Wiley Online Library:** 190+ Journals (Focus: Social Sciences)

These counts reflect the total number of unique journal titles they manage.
```

## Capabilities

### Search academic works
Find specific research papers by title, author, keywords, or date.

### Look up DOI details
Get complete metadata for any published work using its unique digital identifier (DOI).

### Identify journals and publishers
Browse lists of academic journals or find information about specific publishing houses.

### Trace research funding
Discover the organizations that funded a particular piece of research, including work counts and locations.

### Analyze publication types
Filter searches to focus only on specific kinds of academic output, like book chapters or datasets.

## Use Cases

### Verifying a bibliography for publication
A scientific writer needs to confirm every citation in a manuscript. They ask their agent to use `get_work` repeatedly on all DOIs, ensuring every paper has the correct author list and current citation count before submitting.

### Mapping academic relationships
A librarian needs to understand which publishers are associated with specific journals. They run a search using `search_publishers` combined with `search_journals` to map the entire ecosystem of knowledge resources at an institution.

### Investigating research impact
A researcher finds a key paper and needs context. They ask their agent to use `get_funder` on the DOI's metadata, revealing which organizations initially financed that groundbreaking work.

### Broad topic literature review
A student is writing a thesis on AI ethics. Instead of manual searches, they use `search_works` with free-text queries and advanced filters (like publication date or author) to pull a broad set of relevant papers for their bibliography.

## Benefits

- Instantly get complete data on any paper's origin. Instead of guessing, use the `get_work` tool to pull authors, abstract, citation count, and journal info from a single DOI.
- Stop manually cross-referencing journals. Use `search_journals` to find ISSNs and work counts for any academic publication you are researching.
- Track down research funding sources accurately. If you know the topic but not the money trail, use `search_funders` to identify key supporting organizations.
- Build specialized filters into your workflow. Use `get_types` to ensure you only search for specific deliverables, like datasets or book chapters, filtering out noise.
- Batch process metadata retrieval. When reviewing many related papers, pass them all at once using `get_works_by_doi_list` instead of running individual lookups.

## How It Works

The bottom line is, you stop doing manual searches across different academic sites; your agent does it for you.

1. Connect your preferred AI client via Vinkius and enable this MCP.
2. Instruct your agent with a query, such as 'Find all papers on quantum computing from 2023 funded by the DOE.'
3. The MCP processes the request and returns structured data containing titles, authors, DOIs, and full metadata.

## Frequently Asked Questions

**How can CrossRef help me verify citation details for my bibliography?**
CrossRef provides accurate, verifiable metadata for every academic work. By using tools like `get_work`, you get the exact authors, journal name, and up-to-date citation counts, ensuring your bibliography is flawless.

**Do I need to know a DOI to use CrossRef with my AI agent?**
No. While providing a DOI gives you the most precise result, this MCP allows you to search using natural language queries based on keywords, authors, or publication dates.

**Can I find out who paid for a specific piece of research using CrossRef?**
Yes. The MCP includes tools to trace the funding sources. You can discover which organizations funded the work and how many other projects they’ve supported, giving you full context.

**Is this better than just searching Google Scholar for papers?**
This is more powerful because it delivers structured data directly to your agent. Instead of a link list, you get specific fields like ISSNs and work counts that are ready for immediate use in reports.

**How do I find out which types of publications exist (like datasets)?**
You can run a simple query using the `get_types` tool. It gives you a clean, comprehensive list of all recognized publication formats, helping you filter your searches precisely.