核心工具

Configuration File: core_tools.json Tool Type: Local Tools Count: 2

此页面包含在 core_tools.json 配置文件中定义的所有工具。

可用工具

CORE_get_fulltext_snippets (Type: CorePDFSnippetsTool)

Fetch an open-access PDF (commonly returned by CORE_search_papers) and return bounded text snippe…

CORE_get_fulltext_snippets tool specification

工具信息：

  • Name: CORE_get_fulltext_snippets

  • Type: CorePDFSnippetsTool

  • Description: Fetch an open-access PDF (commonly returned by CORE_search_papers) and return bounded text snippets around provided terms. This is a local (download + scan) full-text search fallback when index-based search misses body-only keywords. Extraction backends: PyMuPDF (fitz), pypdf, or markitdown (auto-selects fastest available unless extractor is set).

参数：

  • pdf_url (string) (optional) Direct PDF URL to download (preferred). CORE_search_papers returns this as the url field in many results.

  • url (string) (optional) Alias for pdf_url for convenience when piping CORE_search_papers outputs.

  • terms (array) (required) Terms to search for in the extracted PDF text (case-insensitive).

  • window_chars (integer) (optional) Context window size (characters) before and after each match.

  • max_snippets_per_term (integer) (optional) Maximum number of snippets returned per term.

  • max_total_chars (integer) (optional) Hard cap on total characters returned across all snippets.

  • extractor (string) (optional) PDF text extraction backend: auto (default), fitz (PyMuPDF), pypdf, or markitdown.

  • timeout (integer) (optional) Download timeout in seconds (bounded to <=55 to stay under typical MCP call deadlines).

  • max_pdf_bytes (integer) (optional) Maximum PDF size (bytes) allowed for download/scan. Prevents timeouts on huge PDFs.

  • max_pages (integer) (optional) Maximum number of PDF pages to scan when using fitz/pypdf extractors.

  • max_text_chars (integer) (optional) Maximum number of extracted text characters to scan for term matches (limits CPU/memory).

示例用法：

query = {
    "name": "CORE_get_fulltext_snippets",
    "arguments": {
        "terms": ["item1", "item2"]
    }
}
result = tu.run(query)

**CORE_search_papers**（类型：CoreTool）

Search for open access academic papers using CORE API. CORE aggregates OA repository and journal …

CORE_search_papers 工具规格说明

工具信息：

  • 名称：CORE_search_papers

  • 类型：CoreTool

  • Description: Search for open access academic papers using CORE API. CORE aggregates OA repository and journal content; queries can match metadata and (for many records) indexed full text, but coverage varies by source. Use CORE_get_fulltext_snippets to search inside a returned PDF URL when exact body-text terms matter.

参数：

  • query (string) (optional) Search query for CORE papers. Use keywords separated by spaces to refine your search.

  • search (string) (optional) Alias for query.

  • q (string) (optional) Alias for query.

  • limit (integer) (optional) Maximum number of papers to return. This sets the maximum number of papers retrieved from CORE.

  • page_size (integer) (optional) Alias for limit.

  • max_results (integer) (optional) Alias for limit.

  • ``year_from``（整数）（可选）出版日期筛选的起始年份（例如，2020）。可选参数，用于限制搜索仅包含从该年份起发表的论文。

  • ``year_to``（整数）（可选）出版日期筛选的结束年份（例如，2024年）。可选参数，用于限制搜索范围至该年份及之前发表的论文。

  • ``language``（字符串）（可选）用于筛选论文的语言（例如，’en’、’es’、’fr’）。可选参数，用于限制搜索特定语言的论文。

示例用法：

query = {
    "name": "CORE_search_papers",
    "arguments": {
    }
}
result = tu.run(query)