@shibadityadeb-propacity/propacity-mcp

A production-ready Model Context Protocol (MCP) server that gives any MCP-compatible AI client (Claude Desktop, Cursor, etc.) real-time data about Propacity — India's #1 platform to invest in Bharat's real estate.

What is this?

MCP servers expose tools that AI assistants can call just like functions. Once this server is connected to Claude Desktop, you can ask things like:

• "What does Propacity do?"
• "Who are the founders of Propacity?"
• "Show me the latest news about Propacity."
• "Give me the full Propacity company profile."

…and Claude will call the right tool, fetch live data, and answer you.

Available tools

| Tool | What it returns | |---|---| | get_company_overview | Title, meta description, tagline, headquarters, founded year | | get_social_links | LinkedIn followers, Instagram, Twitter/X, YouTube URLs | | get_funding_and_investors | Total funding raised, rounds, investor list, valuation | | get_team_info | Verified founders with roles, LinkedIn URLs, employee count | | get_latest_news | Recent press articles sorted newest first | | get_full_profile | All of the above in a single merged JSON response |

Install

npm install -g @shibadityadeb-propacity/propacity-mcp

Update to latest version

npm install -g @shibadityadeb-propacity/propacity-mcp@latest

CLI commands

Once installed, use directly from your terminal — no JSON, no pipes:

propacity-mcp overview    # Company title, description, HQ, founded year
propacity-mcp social      # LinkedIn followers, Instagram, Twitter, YouTube
propacity-mcp funding     # Funding rounds, investors, total raised
propacity-mcp team        # Founders with roles + total employee count
propacity-mcp news        # Latest news from Google News RSS
propacity-mcp profile     # Everything above in one call
propacity-mcp --help      # Show all commands

Example output

propacity-mcp social

──────────────────────────────────────────────────
 Social Links
──────────────────────────────────────────────────
 Status : OK
 Fetched: 2025-05-17T06:47:30.497Z
──────────────────────────────────────────────────

 Linkedin:
   Followers: 39106
   Employees: 102
   Location: Delhi
   Url: https://in.linkedin.com/company/propacity
 Instagram:
   Url: https://www.instagram.com/propacity.in/
   Note: Instagram restricts public follower data
 Twitter:
   Url: https://twitter.com/propacity_in
 Youtube:
   Url: https://www.youtube.com/@propacity

Claude Desktop setup

Step 1. Open (or create) the config file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Step 2. Add this and save:

{
  "mcpServers": {
    "propacity": {
      "command": "npx",
      "args": ["-y", "@shibadityadeb-propacity/propacity-mcp"]
    }
  }
}

Step 3. Restart Claude Desktop. Click the hammer icon to confirm all 6 Propacity tools are listed.

If you installed globally, you can also use:

{
  "mcpServers": {
    "propacity": {
      "command": "propacity-mcp"
    }
  }
}

Response format

Every tool returns the same consistent envelope:

{
  "success": true,
  "data": { ... },
  "sources": ["https://propacity.in", "https://..."],
  "fetched_at": "2025-05-17T10:00:00.000Z",
  "errors": []
}

• success — false only if all sources failed
• data — the actual payload (shape varies per tool)
• sources — every URL that was scraped
• fetched_at — ISO timestamp of when the fetch ran
• errors — list of non-fatal partial failures; data is still returned

How it works

Terminal / Claude Desktop
     │
     ▼
src/index.js          ← CLI mode (args) or MCP server (no args, stdio)
     │
     ▼
src/fetchers.js       ← scraping logic (axios + cheerio)
     │
     ├── in.linkedin.com/company/propacity   (followers, employees, location)
     ├── instagram.com/propacity.in          (profile — Meta restricts data)
     ├── news.google.com/rss                 (Google News RSS, no JS needed)
     ├── crunchbase.com                      (funding — 403, uses static data)
     └── yourstory.com                       (funding — 403, uses static data)

Key design decisions:

• Dual mode — pass a command (propacity-mcp overview) for CLI output; no args = MCP stdio server for Claude Desktop
• All HTTP requests have a 12 s timeout, browser-like headers, and legacy TLS support
• get_full_profile fires all 5 fetchers in parallel via Promise.allSettled
• Graceful degradation — blocked sources (Crunchbase/YourStory 403) fall back to known public data with a note

Local development

git clone https://github.com/shibadityadeb-propacity/propacity-mcp.git
cd propacity-mcp
npm install

# Test a single fetcher
node src/index.js overview

# Run as MCP server (for Claude Desktop)
node src/index.js

Project structure

propacity-mcp/
├── src/
│   ├── index.js      # CLI + MCP server entry point
│   └── fetchers.js   # All scraping functions
├── package.json
├── .npmignore
└── README.md

Founders

| Name | Role | LinkedIn | |---|---|---| | Uday Vansh Malik | Co-Founder & CEO | linkedin.com/in/udayvanshmalik | | Rahul Bansal | Co-Founder & CTO | linkedin.com/in/rahulbansal7 | | Imran Shaikh | Co-Founder & COO | linkedin.com/in/imranshaikh4 |

Publishing a new version

npm version patch --no-git-tag-version
npm publish --access public

License

MIT © Propacity