Skip to main content
Glama

Zotero MCP サーバー

MCP(Model Context Protocol)サーバーは、MCPクライアント(Anthropic Claude App、Goose、vscode Clineなど)がローカルのZoteroリポジトリとやり取りできるようにします。このサーバーにより、Zoteroライブラリへのプログラムによるアクセスが可能になり、論文の検索やメモの管理などが可能になります。

設定

  1. 依存関係をインストールします:

pip install -e .
  1. ルート ディレクトリに Zotero の資格情報を含む.envファイルを作成します。

ZOTERO_API_KEY=your_api_key_here
ZOTERO_USER_ID=your_user_id_here

Zotero API キーとユーザー ID は、Zotero の設定ページから取得できます。

Related MCP server: zotero-mcp

Anthropicデスクトップアプリとの統合

Anthropic Desktop アプリと統合するには、 ~/Library/Application Support/Claude/claude_desktop_config.jsonに次の構成を追加します。

{
  "mcpServers": {
    "zotero-mcp-server": {
      "command": "uv",
      "args": [
        "--directory",
        "/Users/swairshah/work/research/zotero-mcp",
        "run",
        "python",
        "-m",
        "zotero_mcp.server"
      ]
    }
  }
}

次のようなエラーが発生した場合

{"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"claude-ai","version":"0.1.0"}},"jsonrpc":"2.0","id":0}
  error: unexpected argument '--directory' found

次に、次の設定を使用し、必ずuv venv ; source .venv/bin/activate ; uv pip install ".[dev]"実行して、サーバーがすべての依存関係で実行できることを確認します。

{
   "mcpServers": {
      "zotero-mcp-server": {
        "command": "bash",
        "args": [
          "-c",
          "cd /Users/shahswai/personal/zotero-mcp-server && source .venv/bin/activate && python -m zotero_mcp.server"
        ]
      }
    }
  }

使用例

Zotero MCP サーバーの例

サーバーでは次のことが可能です。

  • タグで論文を検索

  • 論文の詳細と添付のメモを取得する

  • 論文にメモを追加する

  • 論文概要をリクエストする

Available Tools

10 tools
get_childrenA

Get child items (attachments, notes) of a Zotero item.

Args: key: The Zotero item key.

Returns: JSON array of child items.

Inputs: key (required). Returns: result

get_citationsA

Find papers that cite a given paper using Semantic Scholar.

Args: doi: DOI of the paper. limit: Maximum results (default 100, max 1000). min_citations: Minimum citation count filter. check_library: If true, annotate results with local Zotero presence.

Returns: JSON with count and papers list.

Inputs: doi (required), limit, check_library, min_citations. Returns: result

get_fulltextA

Get full-text content of a Zotero attachment.

Args: key: The key of an attachment item (not a top-level item).

Returns: JSON with content, indexedPages, and totalPages.

Inputs: key (required). Returns: result

get_itemB

Get a single Zotero item by its key.

Args: key: The Zotero item key.

Returns: JSON with the full item data.

Inputs: key (required). Returns: result

get_referencesA

Find papers referenced by a given paper using Semantic Scholar.

Args: doi: DOI of the paper. limit: Maximum results (default 100, max 1000). min_citations: Minimum citation count filter. check_library: If true, annotate results with local Zotero presence.

Returns: JSON with count and papers list.

Inputs: doi (required), limit, check_library, min_citations. Returns: result

list_collectionsA

List all collections in the local Zotero library.

Args: limit: Maximum number of collections to return (0 for all).

Returns: JSON array of collections with id, name, items count, and parent info.

Inputs: limit. Returns: result

list_tagsA

List all tags in the Zotero library, optionally filtered by collection.

Args: collection: Optional collection key to filter tags.

Returns: JSON array of tag strings.

Inputs: collection. Returns: result

search_semantic_scholarA

Search for papers on Semantic Scholar.

Args: query: Search query string. limit: Maximum results (default 20, max 100). year: Year filter, e.g. "2020", "2018-2022", or "2020-". open_access: Only return open access papers. sort: Sort by "citations" or "year" (descending). min_citations: Minimum citation count filter. check_library: If true, annotate results with local Zotero presence.

Returns: JSON with count, total, and papers list.

Inputs: query (required), sort, year, limit, +3 more. Returns: result

Tool Schema Changelog

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

Loading tool schema changelog…

    TDQS

    A3.8/5.0

    Scored across 10 tools

    Disambiguation5/5

    Every tool has a clearly distinct purpose: local Zotero operations (list_collections, list_tags, search, get_item, get_children, get_fulltext) and Semantic Scholar operations (find_related, get_citations, get_references, search_semantic_scholar). Within each group, tools are differentiated by specific actions (e.g., find_related vs. get_citations vs. get_references). There is no ambiguity.

    Naming Consistency5/5

    All tool names follow a consistent verb_noun pattern using lowercase and underscores (e.g., list_collections, get_item, search_semantic_scholar). The verbs appropriately reflect the action: 'list' for enumerations, 'get' for retrieving single items, 'search' for queries, and 'find_related' for semantic similarity. No mixing of conventions.

    Tool Count5/5

    With 10 tools, the surface is well-scoped for a Zotero integration. It covers both local library management (6 tools) and external Semantic Scholar access (4 tools), providing a balanced set without unnecessary overlap or missing core functionality. The count is ideal for an MCP server.

    Completeness3/5

    The server provides robust read operations for Zotero items and comprehensive Semantic Scholar functionality. However, it lacks any write operations (create, update, delete) for Zotero library items or collections, which is a notable gap for a library management server. The tool set appears read-only, limiting its utility for full library management.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers