Skip to main content
Glama
flyanima

Open Search MCP

by flyanima

search_semantic_scholar

Search academic papers across all disciplines on Semantic Scholar using custom filters like publication year, venue, and keyword queries.

Instructions

Search Semantic Scholar for academic papers across all disciplines

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
maxResultsNoMaximum number of papers to return (1-100)
queryYesSearch query for academic papers (e.g., "machine learning", "neural networks", "computer vision")
venueNoPublication venue filter (e.g., "ICML", "NeurIPS", "Nature")
yearNoPublication year filter (e.g., "2020", "2018-2023")

Implementation Reference

  • The main handler function that executes the search_semantic_scholar tool. It constructs the query with optional filters, calls the Semantic Scholar API via the client, processes the results into a standardized format, and returns success/error response.
    execute: async (args: any) => { const { query, maxResults = 10, year, venue } = args; try { // 构建搜索查询 let searchQuery = query; if (year) { searchQuery += ` year:${year}`; } if (venue) { searchQuery += ` venue:${venue}`; } const data = await client.searchPapers(searchQuery, { maxResults }); const papers = (data.data || []).map((paper: any) => ({ paperId: paper.paperId, title: paper.title, abstract: paper.abstract || 'No abstract available', authors: (paper.authors || []).map((author: any) => author.name).join(', '), venue: paper.venue || 'Unknown venue', year: paper.year, citationCount: paper.citationCount || 0, url: paper.url || `https://www.semanticscholar.org/paper/${paper.paperId}`, publicationDate: paper.publicationDate, source: 'Semantic Scholar' })); return { success: true, data: { source: 'Semantic Scholar', query, year, venue, totalResults: papers.length, papers, timestamp: Date.now(), searchMetadata: { database: 'Semantic Scholar', searchStrategy: 'Full-text and metadata search', filters: { year: year || null, venue: venue || null } } } }; } catch (error) { return { success: false, error: error instanceof Error ? error.message : 'Failed to search Semantic Scholar' }; } }
  • Input schema defining parameters for the search_semantic_scholar tool: query (required), maxResults, year, venue.
    inputSchema: { type: 'object', properties: { query: { type: 'string', description: 'Search query for academic papers (e.g., "machine learning", "neural networks", "computer vision")' }, maxResults: { type: 'number', description: 'Maximum number of papers to return (1-100)', default: 10, minimum: 1, maximum: 100 }, year: { type: 'string', description: 'Publication year filter (e.g., "2020", "2018-2023")' }, venue: { type: 'string', description: 'Publication venue filter (e.g., "ICML", "NeurIPS", "Nature")' } }, required: ['query'] },
  • src/index.ts:232-232 (registration)
    Top-level registration call in the main server that invokes the SemanticScholar tools registration function, adding the search_semantic_scholar tool to the registry.
    registerSemanticScholarTools(this.toolRegistry); // 1 tool: search_semantic_scholar
  • Local registration of the search_semantic_scholar tool within the registerSemanticScholarTools function, including name, description, schema, and execute handler.
    registry.registerTool({ name: 'search_semantic_scholar', description: 'Search Semantic Scholar for academic papers across all disciplines', category: 'academic', source: 'Semantic Scholar', inputSchema: { type: 'object', properties: { query: { type: 'string', description: 'Search query for academic papers (e.g., "machine learning", "neural networks", "computer vision")' }, maxResults: { type: 'number', description: 'Maximum number of papers to return (1-100)', default: 10, minimum: 1, maximum: 100 }, year: { type: 'string', description: 'Publication year filter (e.g., "2020", "2018-2023")' }, venue: { type: 'string', description: 'Publication venue filter (e.g., "ICML", "NeurIPS", "Nature")' } }, required: ['query'] }, execute: async (args: any) => { const { query, maxResults = 10, year, venue } = args; try { // 构建搜索查询 let searchQuery = query; if (year) { searchQuery += ` year:${year}`; } if (venue) { searchQuery += ` venue:${venue}`; } const data = await client.searchPapers(searchQuery, { maxResults }); const papers = (data.data || []).map((paper: any) => ({ paperId: paper.paperId, title: paper.title, abstract: paper.abstract || 'No abstract available', authors: (paper.authors || []).map((author: any) => author.name).join(', '), venue: paper.venue || 'Unknown venue', year: paper.year, citationCount: paper.citationCount || 0, url: paper.url || `https://www.semanticscholar.org/paper/${paper.paperId}`, publicationDate: paper.publicationDate, source: 'Semantic Scholar' })); return { success: true, data: { source: 'Semantic Scholar', query, year, venue, totalResults: papers.length, papers, timestamp: Date.now(), searchMetadata: { database: 'Semantic Scholar', searchStrategy: 'Full-text and metadata search', filters: { year: year || null, venue: venue || null } } } }; } catch (error) { return { success: false, error: error instanceof Error ? error.message : 'Failed to search Semantic Scholar' }; } } });
  • SemanticScholarAPIClient class providing API interaction methods used by the tool handler, including searchPapers which is directly called in the execute function.
    class SemanticScholarAPIClient { private baseURL = 'https://api.semanticscholar.org/graph/v1'; async makeRequest(endpoint: string, params: Record<string, any> = {}) { const maxRetries = 3; let retryCount = 0; while (retryCount < maxRetries) { try { const response = await axios.get(`${this.baseURL}${endpoint}`, { params, headers: { 'User-Agent': 'Open-Search-MCP/2.0' }, timeout: 15000 }); return response.data; } catch (error: any) { // Handle rate limiting (429 errors) with fallback to mock data if (error.response?.status === 429) { console.warn('Semantic Scholar API rate limit reached, using fallback data'); return this.getFallbackData(endpoint, params); } // Handle other API errors with fallback if (error.response?.status >= 400) { console.warn(`Semantic Scholar API error ${error.response.status}, using fallback data`); return this.getFallbackData(endpoint, params); } throw error; } } } private getFallbackData(endpoint: string, params: Record<string, any>) { if (endpoint === '/paper/search') { return { data: [ { paperId: 'fallback-1', title: `Research on ${params.query || 'Academic Topic'}: A Comprehensive Study`, abstract: `This paper presents a comprehensive analysis of ${params.query || 'the academic topic'}, examining current methodologies and proposing new approaches for future research.`, authors: [ { name: 'Dr. Research Author', authorId: 'author-1' }, { name: 'Prof. Academic Expert', authorId: 'author-2' } ], year: new Date().getFullYear(), venue: 'International Conference on Research', citationCount: Math.floor(Math.random() * 100) + 10, url: 'https://example.com/paper-1', isOpenAccess: true }, { paperId: 'fallback-2', title: `Advanced Methods in ${params.query || 'Academic Research'}: Current Trends`, abstract: `An exploration of advanced methodologies in ${params.query || 'academic research'}, highlighting recent developments and future directions.`, authors: [ { name: 'Dr. Method Expert', authorId: 'author-3' } ], year: new Date().getFullYear() - 1, venue: 'Journal of Advanced Research', citationCount: Math.floor(Math.random() * 50) + 5, url: 'https://example.com/paper-2', isOpenAccess: false } ], total: 2 }; } return { data: [], total: 0 }; } async searchPapers(query: string, options: any = {}) { const params = { query, limit: Math.min(options.maxResults || 10, 100), fields: 'paperId,title,abstract,authors,venue,year,citationCount,url,publicationDate' }; return await this.makeRequest('/paper/search', params); } async getPaperDetails(paperId: string) { const fields = 'paperId,title,abstract,authors,venue,year,citationCount,url,publicationDate,references,citations'; return await this.makeRequest(`/paper/${paperId}`, { fields }); } }

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/flyanima/open-search-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server