linkup-company-research-mcp 0.2.0

Company research MCP server powered by Linkup - 10 research tools with dual output formats (natural language or structured JSON), full parameter control, and optimized prompts for comprehensive company intelligence.

Linkup Company Research MCP

An MCP server that provides comprehensive company research tools powered by Linkup. Connect it to Claude, Cursor, or any MCP-compatible client to get instant company intelligence.

Features

• 10 research tools covering all aspects of company intelligence
• Dual output formats: Natural language answers with sources, or structured JSON for automation
• Full parameter control: Date filters, domain filters, image support, and result limits
• Optimized prompts: Following Linkup's best practices for accurate, comprehensive research

Tools

Tool	Description	Search Depth
company_overview	Company description, industry, size, business model, products	Deep
company_news	Latest news and developments (with date/topic filters)	Standard
competitive_landscape	Competitors, market position, differentiators	Deep
company_financials	Funding history, valuation, revenue, investors	Deep
company_leadership	Executives, founders, board members, advisors	Standard
company_clients	Customers, case studies, testimonials	Deep
company_technology	Tech stack, patents, engineering, open source	Deep
company_hiring	Job openings, growth signals, Glassdoor ratings	Standard
company_partnerships	Partners, integrations, ecosystem	Deep
company_social_presence	Social media, content strategy, community	Standard

Output Formats

All tools support two output formats via the output_format parameter:

Natural Language (output_format="answer")

Returns a comprehensive answer with up to 5 cited sources. Best for human consumption.

Structured JSON (output_format="structured")

Returns data in a defined JSON schema. Best for automation, CRM integration, and data pipelines.

Parameters

Common Parameters (all tools)

Parameter	Type	Description
company_name	str	Required. The company to research
output_format	str	"answer" (default) or "structured"
max_results	int	Maximum sources to consider (1-50, default: 10-15)

Date Filters (news, financials)

Parameter	Type	Description
from_date	str	Start date in YYYY-MM-DD format
to_date	str	End date in YYYY-MM-DD format

Domain Filters (news)

Parameter	Type	Description
include_domains	str	Comma-separated domains to include (e.g., "techcrunch.com,reuters.com")
exclude_domains	str	Comma-separated domains to exclude

Other Parameters

Parameter	Type	Tools	Description
include_images	bool	overview, competitive, leadership, social	Include relevant images
topic	str	news	Filter by topic (e.g., "funding", "product launch")
department	str	hiring	Filter by department (e.g., "engineering", "sales")

Installation

Claude Desktop

Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "linkup-company-research": {
      "command": "uvx",
      "args": ["linkup-company-research-mcp"],
      "env": {
        "LINKUP_API_KEY": "your-api-key-here"
      }
    }
  }
}

Claude Code

claude mcp add linkup-company-research -- uvx linkup-company-research-mcp

Then set your API key:

export LINKUP_API_KEY="your-api-key-here"

Cursor

Add to .cursor/mcp.json in your project:

{
  "mcpServers": {
    "linkup-company-research": {
      "command": "uvx",
      "args": ["linkup-company-research-mcp"],
      "env": {
        "LINKUP_API_KEY": "your-api-key-here"
      }
    }
  }
}

Get Your API Key

1. Go to linkup.so
2. Sign up or log in
3. Navigate to API settings
4. Generate your API key

Example Usage

Once connected, you can ask your AI assistant things like:

Basic Research

• "Give me an overview of Stripe"
• "What's the latest news about OpenAI?"
• "Who are Figma's main competitors?"

With Parameters

• "Get Anthropic's funding news from the last 6 months" (date filter)
• "Find Notion's customers and case studies" (new tool)
• "What tech stack does Vercel use?" (new tool)

Structured Output for Automation

• "Get Stripe's company overview in structured format" (returns JSON)
• "Research HubSpot's leadership team as JSON" (for CRM integration)

Tool Details

company_overview

Researches the company's website, LinkedIn, and press coverage to provide detailed information about what they do, their industry, size, and business model.

company_news

Searches news sources, press releases, and publications for recent coverage. Supports filtering by date range, topic, and source domains.

competitive_landscape

Identifies competitors, market positioning, differentiators, and competitive advantages through research of industry reports, review sites, and company materials.

company_financials

Researches funding history, valuation, revenue, investors, and financial health through Crunchbase, press releases, and financial news.

company_leadership

Identifies executives, founders, board members, and advisors through company pages, LinkedIn, and press releases.

company_clients

Researches customer pages, case studies, press releases, and review sites to identify verified customers and their use cases.

company_technology

Analyzes engineering blogs, job postings, tech detection tools, patents, and open source contributions to understand technical capabilities.

company_hiring

Researches careers pages, job boards, LinkedIn, and Glassdoor to understand hiring patterns and employee growth signals.

company_partnerships

Researches partner pages, integration marketplaces, press releases, and partner programs to map the company's ecosystem.

company_social_presence

Researches social profiles, content channels, community platforms, and executive thought leadership to understand their digital presence.

Development

# Clone and install locally
git clone https://github.com/LinkupPlatform/linkup-company-research-mcp
cd linkup-company-research-mcp
pip install -e .

# Run the server directly
LINKUP_API_KEY="your-key" linkup-company-research

Project Structure

src/linkup_company_research/
├── __init__.py
├── server.py      # Main MCP server with 10 tools
├── schemas.py     # JSON schemas for structuredOutput
├── prompts.py     # Optimized prompt templates
└── types.py       # Type definitions

License

MIT