bhm-mcp-chat 0.1.0

MCP server for searching BharatMatrimony profiles via GitHub Copilot

BharatMatrimony MCP Search

An MCP (Model Context Protocol) server that lets you search BharatMatrimony profiles directly from VS Code GitHub Copilot Chat. Built with FastMCP.

Ask Copilot things like:

• "Find girls aged 25-30 in Kerala"
• "Search for engineers from Nair caste in Thiruvananthapuram"
• "Show me profiles with Ph.D education who are never married"

The LLM understands all the search filters and translates your natural language queries into structured API calls.

---

Features

Tool	Description
configure_auth	Set your bearer token and session ID (credentials stored locally in .env)
get_auth_status	Check if auth is configured without revealing secrets
search_profiles	Search profiles with 20+ filters — age, height, religion, caste, education, stars (nakshatra), state, occupation, and more

Search Filters

The search_profiles tool exposes all major BharatMatrimony search parameters with rich descriptions so the LLM can use them intelligently:

Parameter	Example Values
age_min / age_max	22, 35
height_min / height_max	FT4_IN7_139CM, FT5_IN9_175CM, FT6_IN0_182CM
religion	HINDU, MUSLIM, CHRISTIAN, SIKH, JAIN, ANY
mother_tongue	MALAYALAM, TAMIL, KANNADA, TELUGU, HINDI, etc.
marital_status	NEVER_MARRIED, DIVORCED, WIDOWED
states	KERALA, KARNATAKA, TAMIL_NADU, MAHARASHTRA, etc.
education_category	BACHELORS_ENGINEERING, MASTERS_MEDICINE_GENERAL, DOCTORATES, etc.
stars (nakshatra)	ROHINI, ASHWINI, REVATHI, SHRAVAN_THIRUVONAM, etc. (all 27)
caste	ANY, NAIR, EZHAVA, NAMBOOTHIRI, PILLAI, etc.
drinking / smoking	NEVER, OCCASIONALLY, NOT_SPECIFIED, ANY
eating	VEGETARIAN, NON_VEGETARIAN, EGGETARIAN, ANY
physical_status	NORMAL, PHYSICALLY_CHALLENGED, ANY
page / results_per_page	Pagination support (default 20 per page)

---

Prerequisites

• Python 3.10+
• A BharatMatrimony account (to obtain API credentials)
• VS Code with GitHub Copilot Chat

---

Installation

Option 1: Clone from GitHub

git clone https://github.com/yourname/bhm-mcp-chat.git
cd bhm-mcp-chat
python -m venv .venv
source .venv/bin/activate   # Linux/Mac
# .venv\Scripts\activate    # Windows
pip install -r requirements.txt

Option 2: Install via pip

pip install bhm-mcp-chat

Option 3: Run directly with uvx (no install needed)

uvx bhm-mcp-chat

---

VS Code Configuration

Add the MCP server to your project's .vscode/mcp.json:

If cloned from GitHub

{
  "servers": {
    "bhm-search": {
      "command": "${workspaceFolder}/.venv/bin/python",
      "args": ["server.py"],
      "cwd": "${workspaceFolder}"
    }
  }
}

If installed via pip

{
  "servers": {
    "bhm-search": {
      "command": "bhm-mcp",
      "args": []
    }
  }
}

If using uvx

{
  "servers": {
    "bhm-search": {
      "command": "uvx",
      "args": ["bhm-mcp-chat"]
    }
  }
}

After adding the config, restart VS Code or reload the window. The MCP server will appear in Copilot's tool list.

---

Getting Your Credentials

The tool needs your BharatMatrimony session credentials. Here's how to get them:

1. Log in to bharatmatrimony.com in your browser
2. Open DevTools (press F12) → go to the Network tab
3. Perform any profile search on the website
4. Find the GraphQL request to g.bharatmatrimony.com in the network log
5. From the Request Headers, copy these values:

Header	What it is
bearer	Your JWT authentication token (long string starting with eyJ...)
sessionId	Your session ID (e.g. 01KMD5TWE3...)
src	Your encrypted user ID (optional, for enc_id)

Note: These tokens expire periodically. You'll need to update them when they expire.

---

Usage

First-time setup (in Copilot Chat)

Just tell Copilot your credentials:

"Configure my BharatMatrimony auth. Bearer token is eyJ... and session ID is 01KMD..."

Copilot will call the configure_auth tool, which saves credentials locally to .env (never committed to git).

Searching profiles

Ask naturally:

"Find girls aged 25-30 in Kerala who are engineers"

"Search for Nair profiles in Thiruvananthapuram with masters degree"

"Show me never-married Hindu profiles from Kerala, page 2"

"Find profiles with compatible nakshatras — Rohini, Revathi, or Thiruvonam"

Copilot translates your query into the right API parameters and returns formatted results with name, age, education, occupation, income, location, caste, and verification status.

Example output

Total: 4,292 matches

1. Aiswarya Nair — 27, 5'3", Ph.D, Professor (₹30L), Thiruvananthapuram, Nair ✅
2. Anju Anand — 28, 5'4", M.Tech, Civil Engineer (₹5L), Kollam, Nair ✅
3. Varsha S — 27, 5'4", B.Tech, Software Professional, Thiruvananthapuram, Ezhava ✅
...

---

Manual setup (alternative to Copilot)

You can also configure credentials by copying the example env file:

cp .env.example .env

Then edit .env with your values:

BHM_USER_ID=E9819315
BHM_ENC_ID=your_enc_id_here
BHM_BEARER_TOKEN=your_bearer_token_here
BHM_SESSION_ID=your_session_id_here

---

Project Structure

bhm-mcp-chat/
├── server.py          # MCP server with 3 tools (configure_auth, get_auth_status, search_profiles)
├── pyproject.toml     # Package config for pip install / PyPI distribution
├── requirements.txt   # Python dependencies
├── .vscode/
│   └── mcp.json       # VS Code MCP server configuration
├── .env.example       # Template for credentials
├── .env               # Your actual credentials (git-ignored)
└── .gitignore

---

How It Works

┌──────────────────┐     stdio/JSON-RPC      ┌─────────────────┐     HTTPS/GraphQL     ┌─────────────────────┐
│  VS Code Copilot │ ◄──────────────────────► │  FastMCP Server │ ──────────────────►   │ BharatMatrimony API │
│  (Chat Agent)    │    MCP Protocol          │  (server.py)    │                       │ g.bharatmatrimony.com│
└──────────────────┘                          └─────────────────┘                       └─────────────────────┘
                                                     │
                                                     ▼
                                                  .env file
                                              (auth credentials)

1. Copilot discovers the MCP tools via .vscode/mcp.json
2. User asks a natural language question
3. Copilot selects the right tool and fills in parameters
4. The server calls BharatMatrimony's GraphQL API
5. Results are returned as structured JSON for Copilot to present

---

Distribution

Method	Command	Best For
GitHub	git clone + pip install -r requirements.txt	Developers, private sharing
PyPI	pip install bhm-mcp-chat	Public distribution
uvx	uvx bhm-mcp-chat (zero-install)	Easiest for end users

To publish to PyPI:

pip install build twine
python -m build
twine upload dist/*

---

Security Notes

• Credentials are stored locally in .env and never committed to git (.gitignore includes .env)
• The get_auth_status tool only reports whether credentials are set — it never exposes token values
• Tokens expire periodically and must be refreshed from the browser

---

License

MIT