Metadata-Version: 2.4
Name: next-gen-code-graph
Version: 1.0.0
Summary: High-performance, multi-language code analysis engine with execution-aware context pruning.
Author: PiQuessen Technologies
License: MIT
Project-URL: Homepage, https://github.com/PiQuessen/next-gen-code-graph
Project-URL: Documentation, https://github.com/PiQuessen/next-gen-code-graph#readme
Keywords: code-graph,static-analysis,mcp,ai-context,tree-sitter,execution-pruning
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Intended Audience :: Developers
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: tree-sitter>=0.22.0
Requires-Dist: tree-sitter-python
Requires-Dist: tree-sitter-kotlin
Requires-Dist: tree-sitter-typescript
Requires-Dist: networkx>=3.1
Requires-Dist: google-genai>=0.4.0
Requires-Dist: pylspclient>=0.1.2
Requires-Dist: python-lsp-server>=1.14.0
Requires-Dist: jedi>=0.19.2
Requires-Dist: click
Requires-Dist: rich
Requires-Dist: mcp
Dynamic: license-file

# 🚀 Next-Gen Code Graph

**Surgical Code Intelligence powered by Execution Pruning.**

Next-Gen Code Graph is a high-performance, multi-language code analysis engine designed to bridge the gap between static AST indexing and dynamic runtime behavior. By combining Tree-sitter's precision with execution coverage analysis, it achieves **8.2x better token efficiency** during code reviews by surgically pruning unexecuted "cold" branches from the context.

---

## ✨ Key Features

- **🌐 23+ Language Support**: Integrated grammars for Python, TypeScript, Kotlin, Go, Rust, C++, and more via a pluggable registry architecture.
- **✂️ Execution Pruning**: Automatically prunes unexecuted code blocks from your analysis context using `coverage.xml` data, saving up to **~74% on LLM tokens**.
- **📓 Jupyter Support**: Direct parsing and indexing of `.ipynb` notebook cells.
- **🔌 One Install, Every Platform**: Auto-configures as an MCP server for Cursor, Claude, Windsurf, Zed, and Continue.
- **🛰️ LSP Integration**: Real-time symbol type resolution via Language Server Protocol.

---

## 🛠️ Installation

```bash
# Clone the repository
git clone https://github.com/PiQuessen/next-gen-code-graph
cd next-gen-code-graph

# Install locally
pip install .

# Auto-configure your AI assistants
next-gen-graph install
```

---

## 🚀 Usage

### Indexing a Workspace
```bash
next-gen-graph index --path ./my-project
```

### Performing a Surgical Review
```bash
# Analyze using execution data for maximum token efficiency
next-gen-graph review --coverage ./coverage.xml
```

---

## 📊 Performance Benchmarks (The PiQue Project)

| Metric | Standard Indexer | Next-Gen Code Graph | Improvement |
| :--- | :--- | :--- | :--- |
| Context Tokens | 120k | 14.5k | **8.2x Reduction** |
| Analysis Speed | ~45s | ~8s | **5.6x Faster** |
| Precision | Generic | Surgical | **High** |

---

## 🛡️ License
Distributed under the MIT License. See `LICENSE` for more information.

---

**Powered by PiQuessen Technologies.**
*"Surgical precision for the modern engineer."*
