Metadata-Version: 2.4
Name: mcp-caste-cypher
Version: 0.1.0
Summary: MCP server for converting caste CSV data to Neo4j Cypher queries using Gemini AI
Author-email: praneeth sunku <praneethsunku@gmail.com>
License: MIT
Keywords: mcp,model-context-protocol,neo4j,cypher,gemini,caste-data,mcp-name: io.github.pareforce/mcp-caste-cypher
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.12
Classifier: Intended Audience :: Developers
Classifier: Topic :: Software Development :: Libraries
Classifier: License :: OSI Approved :: MIT License
Requires-Python: >=3.12
Description-Content-Type: text/markdown
Requires-Dist: mcp>=1.14.1
Requires-Dist: httpx>=0.27.0
Requires-Dist: google-genai>=1.0.0

# MCP Caste Cypher Generator

A Model Context Protocol (MCP) server that converts caste CSV data into Neo4j Cypher queries using Google's Gemini AI batch processing.

mcp-name: io.github.pareforce/mcp-caste-cypher

## Features

- **CSV to JSONL Conversion**: Prepare batch requests from CSV files containing caste information
- **Batch Processing**: Leverage Gemini AI's batch API for efficient large-scale processing
- **Cypher Query Generation**: Automatically generate Neo4j graph database queries
- **Job Monitoring**: Track batch job status in real-time

## Tools Provided

### 1. `prepare_batch_input`
Converts CSV files into JSONL batch request format.

**Parameters:**
- `input_csv`: Path to input CSV file
- `output_jsonl`: Path to output JSONL file

### 2. `run_batch_job`
Executes the complete batch workflow: upload, process, monitor, and download results.

**Parameters:**
- `api_key`: Google Gemini API key
- `input_jsonl`: Path to input JSONL file
- `output_cypher`: Path to output Cypher file (automatically adds .cypher extension)

### 3. `check_batch_status`
Monitor the status of a running batch job.

**Parameters:**
- `api_key`: Google Gemini API key
- `job_name`: Name of the batch job to check

## Installation

```bash
pip install mcp-caste-cypher
```

## Usage

### With Claude Desktop

Add to your Claude Desktop configuration:

```json
{
  "mcpServers": {
    "caste-cypher": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/mcp-caste-cypher",
        "run",
        "mcp-caste-cypher"
      ]
    }
  }
}
```

### Standalone

```bash
uv run mcp-caste-cypher
```

## Supported CSV Formats

**Format 1 (BC/SC with groups):**
```
serial_number,group_name,caste_name,category
1,BC-A,Agnikulakshatriya,BC
```

**Format 2 (General/ST):**
```
serial_number,caste_name,category
1,Kapu,GENERAL
```

## Requirements

- Python >=3.12
- Google Gemini API key
- Dependencies: `mcp`, `httpx`, `google-genai`

## License

MIT

## Author

Praneeth Sunku (praneethsunku@gmail.com)
