Job Search MCP Server

Instructions

Implementation Reference

• src/index.ts:160-184 (handler)

  The handler function for the 'search_specific_job_site' tool. It parses arguments, builds a JobFilter for AI/ML entry-level jobs requiring Python, calls jobSearchService.searchSite, and returns the results as a JSON string in the MCP response format.

  private async searchSpecificJobSite(args: any) {
    const { site, location = 'Remote', maxResults = 25 } = args;
  
    const filter: JobFilter = {
      location,
      maxResults,
      includeInternships: true,
      includeFullTime: true,
      keywords: [],
      experienceLevel: 'entry',
      requiredSkills: ['python'],
      jobTypes: ['ai', 'ml', 'machine learning', 'artificial intelligence', 'data science']
    };
  
    const results = await this.jobSearchService.searchSite(site, filter);
    
    return {
      content: [
        {
          type: 'text',
          text: JSON.stringify(results, null, 2)
        }
      ]
    };
  }

• src/index.ts:76-96 (schema)

  Input schema defining parameters for the tool: site (required, enum of job sites), location, maxResults.

  inputSchema: {
    type: 'object',
    properties: {
      site: {
        type: 'string',
        enum: ['linkedin', 'indeed', 'glassdoor', 'ziprecruiter', 'monster'],
        description: 'Job site to search'
      },
      location: {
        type: 'string',
        description: 'Job location',
        default: 'Remote'
      },
      maxResults: {
        type: 'number',
        description: 'Maximum number of results to return',
        default: 25
      }
    },
    required: ['site']
  }

• src/index.ts:107-110 (registration)

  Registration in the tool dispatch switch statement within the CallToolRequestSchema handler.

  case 'search_ai_ml_jobs':
    return await this.searchAIMLJobs(args);
  case 'search_specific_job_site':
    return await this.searchSpecificJobSite(args);

• src/index.ts:73-97 (registration)

  Tool registration in the ListToolsRequestSchema response, including name, description, and input schema.

  {
    name: 'search_specific_job_site',
    description: 'Search for jobs on a specific job site',
    inputSchema: {
      type: 'object',
      properties: {
        site: {
          type: 'string',
          enum: ['linkedin', 'indeed', 'glassdoor', 'ziprecruiter', 'monster'],
          description: 'Job site to search'
        },
        location: {
          type: 'string',
          description: 'Job location',
          default: 'Remote'
        },
        maxResults: {
          type: 'number',
          description: 'Maximum number of results to return',
          default: 25
        }
      },
      required: ['site']
    }
  }

• src/services/jobSearchService.ts:52-79 (helper)

  Core helper method called by the handler. Retrieves the site-specific scraper, calls its search method with the filter, applies post-search filtering, and returns structured results or error.

  async searchSite(site: string, filter: JobFilter): Promise<JobSearchResult> {
    const startTime = Date.now();
    const scraper = this.scrapers.get(site.toLowerCase());
    
    if (!scraper) {
      throw new Error(`Unsupported job site: ${site}`);
    }
  
    try {
      const jobs = await scraper.search(filter);
      const filteredJobs = this.filterJobs(jobs, filter);
      
      return {
        site,
        jobs: filteredJobs,
        totalFound: filteredJobs.length,
        searchTime: Date.now() - startTime
      };
    } catch (error) {
      return {
        site,
        jobs: [],
        totalFound: 0,
        searchTime: Date.now() - startTime,
        error: error instanceof Error ? error.message : 'Unknown error'
      };
    }
  }