Skip to main content
Glama
deenesjakoruzh
tfscharff

DOI Citation Verifier

by tfscharff
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Remote

🚀 Быстрая установка

npx -y github:tfscharff/doi-mcp

Или добавьте в конфигурацию Claude Desktop:

{
  "mcpServers": {
    "doi-mcp": {
      "command": "npx",
      "args": ["-y", "github:tfscharff/doi-mcp"]
    }
  }
}

Related MCP server: CiteStamp MCP server

Какую проблему это решает

Большие языковые модели иногда «галлюцинируют» академические цитаты — ссылаются на несуществующие статьи, приписывают реальные названия не тем авторам или путают детали публикации. Этот MCP-сервер устраняет данную проблему за счет:

  1. Проверки по 9 базам данных: Проверяет цитаты в CrossRef, OpenAlex, PubMed, zbMATH, ERIC, HAL, INSPIRE-HEP, Semantic Scholar и DBLP

  2. Параллельного поиска: Запрашивает все базы данных одновременно для получения быстрых результатов (~1 секунда)

  3. Всестороннего охвата: Более 600 миллионов публикаций по всем дисциплинам, включая STEM, гуманитарные, социальные науки и образование

  4. Цитат с поддержкой DOI: Каждая проверенная цитата включает действительный, кликабельный DOI

Функции

  • Поиск по 9 базам данных: CrossRef, OpenAlex, PubMed, zbMATH, ERIC, HAL, INSPIRE-HEP, Semantic Scholar, DBLP

  • Верификация цитат: Проверка того, существует ли статья с конкретными данными во всех базах

  • Поиск проверенных статей: Поиск реальных статей по теме с получением только проверенных цитат

  • Параллельная обработка: Все запросы к базам данных выполняются одновременно для максимальной скорости

  • Оптимизация производительности: Умное кэширование и стратегии раннего завершения для ускорения проверки на 25-35%

  • Выбор источников: Поиск по всем базам данных или по конкретным источникам

  • Форматирование цитат: Возвращает правильно оформленные цитаты с DOI

  • Нулевая конфигурация: Все базы данных работают «из коробки» без необходимости использования API-ключей

Как это работает

Когда ИИ-ассистента спрашивают об исследованиях или цитатах:

  1. Без этого MCP: Ассистент может сослаться на «Согласно Smith et al. (2023) в Nature...», ссылаясь на статью, которой не существует

  2. С этим MCP: Ассистент сначала использует verifyCitation, который выполняет поиск по 9 базам данных параллельно и возвращает:

    • Проверенное совпадение с полным DOI → Можно цитировать

    • Совпадение не найдено → Нельзя цитировать; необходимо искать реальные статьи

Инструменты

verifyCitation

Основной инструмент против галлюцинаций — проверяет существование цитаты по нескольким базам данных перед тем, как она будет упомянута.

Входные данные:

  • title (строка, необязательно): Название статьи (принимаются частичные совпадения)

  • authors (массив, необязательно): Имена авторов (достаточно фамилий)

  • year (число, необязательно): Год публикации

  • doi (строка, необязательно): DOI, если известен

  • journal (строка, необязательно): Название журнала

Возвращает JSON с:

  • verified: true/false

  • Если verified=true: DOI, название, авторы, год, журнал, URL, исходная база данных

  • Если verified=false: Предупреждающее сообщение о том, что соответствующая публикация не найдена

  • Индикаторы качества совпадения для прозрачности

Пример успешной верификации:

{
  "verified": true,
  "doi": "10.1038/s41586-023-06004-9",
  "title": "Accurate structure prediction of biomolecular interactions...",
  "authors": ["John Jumper", "Richard Evans", "..."],
  "year": 2023,
  "journal": "Nature",
  "url": "https://doi.org/10.1038/s41586-023-06004-9",
  "source": "crossref",
  "message": "✓ Citation verified"
}

findVerifiedPapers

Поиск реальных статей по теме и возврат только проверенных цитат с DOI из нескольких баз данных.

Входные данные:

  • query (строка): Поисковый запрос (тема, ключевые слова, имена авторов)

  • source (строка, необязательно): В какой базе данных искать — "all" (по умолчанию), "crossref", "openalex", "pubmed", "zbmath", "eric", "hal", "inspirehep", "semanticscholar" или "dblp"

  • limit (число, необязательно): Количество результатов на источник (1-20, по умолчанию: 5)

  • yearFrom (число, необязательно): Минимальный год публикации

  • yearTo (число, необязательно): Максимальный год публикации

Возвращает: Массив проверенных статей из указанной базы данных (баз данных) с полной информацией о цитировании, включая источник

Пример:

// Search all 9 databases
findVerifiedPapers({ query: "CRISPR gene editing", limit: 5 })

// Search only PubMed for biomedical papers
findVerifiedPapers({ query: "cancer immunotherapy", source: "pubmed", limit: 10 })

// Search zbMATH for mathematics papers
findVerifiedPapers({ query: "algebraic topology", source: "zbmath" })

// Search DBLP for computer science papers
findVerifiedPapers({ query: "neural networks", source: "dblp", yearFrom: 2020 })

// Search ERIC for education research
findVerifiedPapers({ query: "active learning pedagogy", source: "eric" })

// Search HAL for French/European humanities research
findVerifiedPapers({ query: "phenomenology Husserl", source: "hal" })

// Search INSPIRE-HEP for high-energy physics papers
findVerifiedPapers({ query: "Higgs boson", source: "inspirehep" })

Установка

Добавьте в файл конфигурации Claude Desktop:

Windows: %APPDATA%\Claude\claude_desktop_config.json macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Linux: ~/.config/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "doi-mcp": {
      "command": "npx",
      "args": ["-y", "github:tfscharff/doi-mcp"]
    }
  }
}

Перезапустите Claude Desktop, и сервер станет доступен.

Альтернатива: Глобальная установка

npm install -g github:tfscharff/doi-mcp

Затем используйте эту конфигурацию:

{
  "mcpServers": {
    "doi-mcp": {
      "command": "doi-mcp"
    }
  }
}

Альтернатива: Локальное клонирование

git clone https://github.com/tfscharff/doi-mcp.git
cd doi-mcp
npm install
npm run build

Конфигурация для локальной установки:

{
  "mcpServers": {
    "doi-mcp": {
      "command": "node",
      "args": ["/absolute/path/to/doi-mcp/dist/index.js"]
    }
  }
}

Устранение неполадок

Сервер не подключается

  1. Проверьте, установлен ли Node.js: node --version (требуется v18+)

  2. Проверьте логи Claude Desktop:

    • Windows: %APPDATA%\Claude\logs\

    • macOS: ~/Library/Logs/Claude/

    • Linux: ~/.config/Claude/logs/

Ошибка команды npx

npm cache clean --force

Локальное тестирование

npx @modelcontextprotocol/inspector node dist/index.js

Разработка

# Install dependencies
npm install

# Build
npm run build

# Development with watch mode
npm run dev

Пример использования

До этого MCP (галлюцинации при цитировании):

User: "Tell me about recent AlphaFold research"
Assistant: "According to Johnson et al. (2024) in Science, AlphaFold3 achieved..."
           ❌ This paper doesn't exist

После этого MCP (только проверенные цитаты):

User: "Tell me about recent AlphaFold research"
Assistant: [Uses findVerifiedPapers tool]
           "According to Jumper et al. (2023) in Nature (DOI: 10.1038/s41586-023-06004-9), 
            AlphaFold3 achieved..."
           ✓ Real paper with valid DOI verified across databases

Верификация выявляет фальшивые цитаты:

User: "Can you verify this citation: Smith et al. (2024), 'Quantum AI', Nature"
Assistant: [Uses verifyCitation tool - searches all 9 databases in parallel]
           "⚠ I cannot verify this citation - no matching publication found in
            any of the 9 databases. This citation may be incorrect."

Охват баз данных

Все базы данных опрашиваются параллельно для максимальной скорости (~1 секунда в сумме):

Общие базы данных

  • CrossRef: 150+ миллионов научных публикаций по всем дисциплинам

  • OpenAlex: 250+ миллионов научных работ по всем дисциплинам

  • Semantic Scholar: 200+ миллионов статей с поиском на базе ИИ

Специализированные базы данных

  • PubMed: 35+ миллионов публикаций по биомедицине и наукам о жизни

  • zbMATH: 4+ миллиона публикаций по математике

  • DBLP: Исчерпывающая библиография по информатике (журналы и конференции)

  • ERIC: 1.7+ миллиона публикаций по исследованиям в области образования

  • HAL: 4.4+ миллиона французских/европейских научных документов (2.5 млн на английском)

  • INSPIRE-HEP: 1.7+ миллиона публикаций по физике высоких энергий

Общий охват

600+ миллионов публикаций по всем академическим дисциплинам со специализированной глубиной в STEM, информатике, биомедицинских науках, математике и исследованиях в области образования.

Лицензия

MIT

Участие в проекте

Вклад приветствуется! Пожалуйста, не стесняйтесь присылать отчеты об ошибках или pull-запросы.

Связанные ресурсы

Документация API

Ресурсы

Available Tools

3 tools
batchVerifyCitationsA
Read-onlyIdempotent

Verify multiple citations in a single call. More efficient than calling verifyCitation multiple times. Returns verification status for each citation.

ParametersJSON Schema
NameRequiredDescriptionDefault
citationsYesArray of citations to verify

TDQS

A4.4/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description adds context beyond annotations by specifying that it 'Returns verification status for each citation,' which clarifies the output behavior. Annotations already indicate it's read-only, idempotent, and non-destructive, so the description doesn't need to repeat those traits, but it usefully describes the return format.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise and front-loaded, consisting of two sentences that efficiently convey the tool's purpose, efficiency benefit, and return value without any wasted words. Every sentence adds value, making it easy for an agent to parse quickly.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's moderate complexity, the description is complete enough: it covers purpose, usage guidelines, and output behavior. With annotations handling safety traits and no output schema, the description fills gaps by explaining the return format. However, it could briefly mention error handling or limits for full completeness.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The description mentions 'citations' as the input but doesn't add semantic details beyond what the schema provides. With 100% schema description coverage, the schema fully documents the 'citations' array and its nested properties, so the baseline score of 3 is appropriate as the description doesn't compensate with extra parameter insights.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose with a specific verb ('Verify multiple citations') and resource ('citations'), distinguishing it from sibling tools like 'verifyCitation' by emphasizing batch processing efficiency. It explicitly mentions the return value ('verification status for each citation'), which adds clarity.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides explicit guidance on when to use this tool versus alternatives: it states 'More efficient than calling verifyCitation multiple times,' directly comparing it to a sibling tool. This helps the agent choose this tool for batch operations over single-citation verification.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

findVerifiedPapersA
Read-onlyIdempotent

Search multiple academic databases (CrossRef, OpenAlex, PubMed, zbMATH, ERIC, HAL, INSPIRE-HEP, Semantic Scholar, DBLP) for papers and return only verified, real citations with DOIs.

ParametersJSON Schema
NameRequiredDescriptionDefault
queryYesSearch query (topic, keywords, author names)
limitNoNumber of results per source
yearFromNoMinimum publication year
yearToNoMaximum publication year
sourceNoWhich source to searchall

TDQS

A4.2/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already indicate read-only, idempotent, and non-destructive behavior. The description adds valuable context beyond this: it specifies the multiple databases searched (CrossRef, OpenAlex, etc.) and the verification requirement (only papers with DOIs are returned). This helps the agent understand the tool's scope and output quality, though it doesn't mention rate limits or authentication needs.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, dense sentence that efficiently conveys the tool's purpose, scope, and key behavior. It lists all databases upfront and specifies the verification requirement without unnecessary words. Every element earns its place, making it highly concise and well-structured.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (searching multiple databases with verification), annotations cover safety (read-only, non-destructive), and schema fully documents parameters, the description provides good contextual completeness. It explains the multi-source approach and DOI verification, though without an output schema, it doesn't detail the return format (e.g., what fields are included).

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, providing full parameter documentation. The description doesn't add any parameter-specific details beyond what's in the schema (e.g., it doesn't explain query syntax or source differences). With high schema coverage, the baseline score of 3 is appropriate as the description doesn't compensate but doesn't need to.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('search multiple academic databases'), the resource ('papers'), and a key distinguishing feature ('return only verified, real citations with DOIs'). It differentiates from siblings by focusing on multi-source search with verification, unlike batchVerifyCitations and verifyCitation which likely handle verification of existing citations rather than searching.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage context by specifying it searches 'multiple academic databases' and returns 'verified, real citations with DOIs', suggesting it's for finding reliable academic sources. However, it doesn't explicitly state when to use this tool versus its siblings (batchVerifyCitations, verifyCitation), which likely handle different verification scenarios.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

verifyCitationA
Read-onlyIdempotent

CRITICAL: Use this to verify ANY academic citation before mentioning it. Checks multiple databases (CrossRef, OpenAlex, PubMed, zbMATH, ERIC, HAL, INSPIRE-HEP, Semantic Scholar, DBLP) if a paper exists. Returns null if not found.

ParametersJSON Schema
NameRequiredDescriptionDefault
titleNoPaper title (partial matches accepted)
authorsNoAuthor names (last names sufficient)
yearNoPublication year
doiNoDOI if known
journalNoJournal name

TDQS

A4.4/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description adds valuable behavioral context beyond annotations: it lists the specific databases checked (CrossRef, OpenAlex, etc.) and states that it 'returns null if not found,' which clarifies the output behavior. Annotations already indicate it's read-only, idempotent, and non-destructive, so the description doesn't need to repeat those traits, but it enhances understanding with operational details.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is highly concise and well-structured: it starts with a critical warning, states the purpose and usage in a single sentence, lists databases efficiently, and ends with return behavior. Every sentence adds essential information without redundancy, making it front-loaded and easy to parse.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (verifying citations across multiple databases) and the absence of an output schema, the description is mostly complete: it explains the purpose, usage, databases checked, and return behavior. However, it lacks details on error handling, rate limits, or authentication needs, which could be useful for full transparency. The annotations cover safety aspects, so it's adequate but not exhaustive.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

With 100% schema description coverage, the input schema fully documents all 5 parameters (title, authors, year, doi, journal), including details like 'partial matches accepted' for title and 'last names sufficient' for authors. The description adds no additional parameter information, so it meets the baseline of 3 by not duplicating schema content.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose with a specific verb ('verify') and resource ('academic citation'), explicitly distinguishes it from siblings by specifying it's for verifying citations before mentioning them (unlike batchVerifyCitations or findVerifiedPapers), and provides critical context about checking multiple databases. The 'CRITICAL' prefix emphasizes its importance.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description explicitly states when to use this tool ('before mentioning [a citation]') and provides clear alternatives by naming sibling tools (batchVerifyCitations, findVerifiedPapers), though it doesn't detail when to use those instead. The 'CRITICAL' label implies it should be used for any citation verification, making the guidance comprehensive.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

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

  1. 3 tool updatesv1.0.0
    • First observedbatchVerifyCitations
    • First observedfindVerifiedPapers
    • First observedverifyCitation
Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A4.3/5.0

Scored across 3 tools

Disambiguation5/5

Each tool has a clearly distinct purpose: batchVerifyCitations handles multiple citations efficiently, verifyCitation checks individual citations, and findVerifiedPapers searches databases for verified papers. There is no overlap in functionality, making tool selection straightforward for an agent.

Naming Consistency4/5

The naming follows a consistent verb_noun pattern (batchVerifyCitations, findVerifiedPapers, verifyCitation), with all tools using camelCase. However, verifyCitation lacks a noun suffix like 'Citation' in its verb part, which is a minor deviation from perfect consistency.

Tool Count4/5

With 3 tools, the count is reasonable for a DOI citation verification server, covering core operations (verify single, verify batch, search verified). It might be slightly thin, as additional tools for managing results or databases could enhance completeness, but it's well-scoped for the basic purpose.

Completeness4/5

The tool set covers key verification tasks: single and batch verification, plus searching for verified papers. Minor gaps exist, such as tools for updating or deleting verification data, but the core workflow of verifying and finding citations is adequately supported without dead ends.

Maintenance

ActivityStale
ResponsivenessUnresponsive

Resources

Related MCP Connectors

Related MCP Servers

  • CiteStamp MCP server by citestamp

    CiteStamp MCP serverofficial

    citestamp
    A
    license
    Not graded
    quality
    B
    maintenance
    Ground citations before your agent emits them by checking references against public scholarly registries and flagging hallucinated or retracted ones.
    MIT

  • AgentCite

    Kevinx702
    F
    license
    Not graded
    quality
    D
    maintenance
    Fabrication-free, DOI-backed citations for AI content and agents, using openAlex public-domain data with resolvable DOIs. Includes an API and planned MCP server for agent-native citation retrieval.
    -