# Context Graph Miner

## Overview

`context_graph_miner` is a CLI tool for mining patterns and insights from context graphs generated by AI agents. It enables developers to identify decision-making trends, bottlenecks, and areas for improvement by analyzing the stored graph data.

## Features

- Load context graphs from JSON files.
- Query nodes in the graph based on attributes.
- Generate reports summarizing nodes and edges in CSV or JSON formats.

## Installation

Install the required dependencies:

```bash
pip install networkx pandas
```

## Usage

Run the tool using the command line:

```bash
python context_graph_miner.py --file <path_to_graph.json> [--query <query_string>] [--report <csv|json>]
```

### Arguments

- `--file`: Path to the JSON file containing the context graph (required).
- `--query`: Query string to search for specific nodes (optional).
- `--report`: Generate a report in the specified format (`csv` or `json`) (optional).

### Example

```bash
python context_graph_miner.py --file graph.json --query decision --report csv
```

## Testing

Run the tests using `pytest`:

```bash
pytest test_context_graph_miner.py
```

## License

MIT License