neurocode-ai 0.1.0__tar.gz → 1.2.0__tar.gz

{neurocode_ai-0.1.0/neurocode_ai.egg-info → neurocode_ai-1.2.0}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: neurocode-ai
-Version: 0.1.0
-Summary: Heterogeneous Code-Graph Slicing for On-Device Autonomous SWE Agents
+Version: 1.2.0
+Summary: Heterogeneous Code-Graph Slicing & AST Subtree Grafting for On-Device Autonomous SWE Agents
 Author: Deepak R.
 License: MIT
 Requires-Python: >=3.10
@@ -9,6 +9,8 @@ Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: tree-sitter>=0.22.0
 Requires-Dist: tree-sitter-python>=0.21.0
+Requires-Dist: tree-sitter-javascript>=0.21.0
+Requires-Dist: tree-sitter-typescript>=0.21.0
 Requires-Dist: networkx>=3.0
 Requires-Dist: datasets>=2.18.0
 Requires-Dist: matplotlib>=3.8.0
@@ -17,12 +19,14 @@ Dynamic: license-file
 
 # NeuroCode: Heterogeneous Code-Graph Slicing for On-Device Autonomous Software Engineering
 
+[![PyPI version](https://img.shields.io/pypi/v/neurocode-ai.svg)](https://pypi.org/project/neurocode-ai/)
+[![Visual Studio Marketplace](https://img.shields.io/visual-studio-marketplace/v/DeepakR.neurocode-ai.svg?label=VS%20Code%20Marketplace)](https://marketplace.visualstudio.com/items?itemName=DeepakR.neurocode-ai)
+[![Glama MCP Server](https://glama.ai/mcp/servers/Arkz-Deepak/neurocode/badge)](https://glama.ai/mcp/servers/Arkz-Deepak/neurocode)
+[![Awesome MCP Servers](https://img.shields.io/badge/Awesome--MCP--Servers-Listed-green.svg)](https://github.com/punkpeye/awesome-mcp-servers)
+[![ICPC 2027](https://img.shields.io/badge/ICPC%202027-Submission%20%2363-purple.svg)](https://conf.researchr.org/home/icpc-2027)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 [![Python 3.10+](https://img.shields.io/badge/python-3.10+-blue.svg)](https://www.python.org/downloads/)
 [![Hardware: 4GB GPU](https://img.shields.io/badge/Hardware-4GB%20VRAM%20Verified-brightgreen.svg)]()
-[![Model Context Protocol](https://img.shields.io/badge/MCP-2024--11--05%20Stdio-blueviolet.svg)](https://modelcontextprotocol.io)
-[![Ollama Integration](https://img.shields.io/badge/Ollama-Offline%20Native-orange.svg)](https://ollama.ai)
-[![Kaggle: Verified](https://img.shields.io/badge/Kaggle-Notebook%20Live-20beff.svg)](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
 
 **NeuroCode** is a lightweight, heterogeneous neuro-symbolic framework and **Model Context Protocol (MCP)** server that enables small, local language models (1.5B–14B parameters) to fix multi-file software engineering bugs on consumer laptops (e.g., 4GB–8GB VRAM) without cloud API dependency or Out-of-Memory (OOM) crashes.
 
@@ -51,11 +55,11 @@ NeuroCode was evaluated across 5 comprehensive benchmark tracks spanning real-wo
 
 | Benchmark Suite | Metric | Baseline / BM25 | NeuroCode (Ours) | Advantage |
 | :--- | :--- | :--- | :--- | :--- |
-| **SWE-bench Lite** ($N=300$) | Resolve Rate (Pass@1) | 0.0% (OOM on 4GB) | **41.4%** (Local) / **21.1%** (14B) | Prevents OOM cliff; +14.2% over BM25 |
-| **HumanEval** ($N=164$) | Pass@1 | 80.5% | **86.0%** (141/164) | +5.5% via syntax-aware graph repair |
-| **MBPP** ($N=100$) | Pass@1 | 81.0% | **87.0%** (87/100) | +6.0% algorithmic verification |
+| **SWE-bench Lite** ($N=292$) | Valid Diff Apply / Resolved | 0.0% (OOM on 4GB) / 0.0% | **41.4% apply / 9.5% resolved** | Prevents OOM cliff; +23.9% apply over BM25 |
+| **HumanEval** ($N=164$) | Pass@1 (Unit Test) | 80.5% | **86.0%** (141/164) | +5.5% via syntax-aware graph repair |
+| **MBPP** ($N=100$) | Pass@1 (Unit Test) | 81.0% | **87.0%** (87/100) | +6.0% algorithmic verification |
 | **RepoBench-R** ($N=150$) | Recall@1 / Token Budget | 18.0% (745 tokens) | **100.0%** (**428 tokens**) | **-42.6% tokens**, perfect cross-file retrieval |
-| **Host Systems Telemetry** | Active VRAM / Slice Latency | 14.2 GB (OOM) | **961.2 MB** / **1.09 ms** | **93.2% VRAM reduction**, 0 OOMs |
+| **Host Systems Telemetry** | Active VRAM / Slice Latency | 14.2 GB (OOM) | **961.2 MB** / **0.62 ms** | **93.2% VRAM reduction**, 0 OOMs |
 
 ### 💻 Physical Laptop Edge Validation (NVIDIA RTX 3050 Laptop GPU)
 Tested live on an everyday developer laptop (AMD Ryzen 7 7435HS + NVIDIA RTX 3050 Laptop GPU, 4.0 GB VRAM):
@@ -124,16 +128,31 @@ Add to your `~/.continue/config.json`:
 
 ---
 
+## 💻 Visual Studio Code Extension
+
+Install directly from the [Visual Studio Marketplace](https://marketplace.visualstudio.com/items?itemName=DeepakR.neurocode-ai):
+
+```bash
+code --install-extension DeepakR.neurocode-ai
+```
+
+*(Or download the `.vsix` binary from the [v0.1.0 Release](https://github.com/Arkz-Deepak/neurocode/releases/tag/v0.1.0))*
+
+### Extension Features:
+- **Status Bar Telemetry**: Live VRAM headroom and `/dev/shm` RAM health indicators.
+- **Context Slicing**: Right-click any active file or code block to extract sub-1,200 token reachability closure in $<100$ ms.
+- **1-Click Prompt Copy**: Instant paste into Cursor, Claude, or Copilot without context overflow.
+- **Autonomous Diff Viewer**: Direct side-by-side patch inspection powered by local Ollama backbones.
+
+---
+
 ## 🚀 Quickstart & CLI Usage
 
-### 1. Installation
+### 1. Installation via PyPI
 ```bash
-git clone https://github.com/Arkz-Deepak/neurocode.git
-cd neurocode
-python3 -m venv .venv
-source .venv/bin/activate
-pip install -e .
+pip install neurocode-ai
 ```
+*(Or install locally from source: `git clone https://github.com/Arkz-Deepak/neurocode.git && cd neurocode && pip install -e .`)*
 
 ### 2. Extract Token-Bounded Slice for Any Issue
 ```bash
@@ -170,10 +189,15 @@ neurocode-copilot
 
 ## 📄 Publications & Preprints
 
-- **IEEE Transactions on Software Engineering (TSE)**: [`paper_ieee_tse/neurocode_tse.tex`](paper_ieee_tse/neurocode_tse.tex)
-- **IEEE Conference Modular Package**: [`paper_modular_ieee/neurocode_modular_paper.tex`](paper_modular_ieee/neurocode_modular_paper.tex)
-- **MLSys Conference Format**: [`paper_mlsys/neurocode_mlsys.tex`](paper_mlsys/neurocode_mlsys.tex)
-- **Kaggle Competition Executable Notebook**: [Kaggle Notebook Live](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
+- **35th IEEE/ACM International Conference on Program Comprehension (ICPC 2027)**:
+  - *Track*: Tool Demonstration Track (Submission #63)
+  - *Title*: *NeuroCode: In-Memory Code Property Graph Slicing and Edge SWE Agent for Resource-Constrained Environments*
+  - *Source*: [`paper_icpc_tool_demo/neurocode_icpc_demo.tex`](paper_icpc_tool_demo/neurocode_icpc_demo.tex)
+- **Journal of Systems and Software (Elsevier)**:
+  - *Track*: Full Research Article (Traditional Subscription, ₹0 APC)
+  - *Title*: *On-Device Autonomous Software Engineering: In-Memory Code Property Graph Reachability Slicing under Edge VRAM Constraints*
+  - *Source*: [`paper_elsevier_jss/neurocode_jss.tex`](paper_elsevier_jss/neurocode_jss.tex)
+- **Kaggle Empirical Reproduction**: [Kaggle Notebook Live](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
 
 ### Citation
 ```bibtex

neurocode_ai-0.1.0/PKG-INFO → neurocode_ai-1.2.0/README.md

@@ -1,28 +1,13 @@
-Metadata-Version: 2.4
-Name: neurocode-ai
-Version: 0.1.0
-Summary: Heterogeneous Code-Graph Slicing for On-Device Autonomous SWE Agents
-Author: Deepak R.
-License: MIT
-Requires-Python: >=3.10
-Description-Content-Type: text/markdown
-License-File: LICENSE
-Requires-Dist: tree-sitter>=0.22.0
-Requires-Dist: tree-sitter-python>=0.21.0
-Requires-Dist: networkx>=3.0
-Requires-Dist: datasets>=2.18.0
-Requires-Dist: matplotlib>=3.8.0
-Requires-Dist: transformers>=4.40.0
-Dynamic: license-file
-
 # NeuroCode: Heterogeneous Code-Graph Slicing for On-Device Autonomous Software Engineering
 
+[![PyPI version](https://img.shields.io/pypi/v/neurocode-ai.svg)](https://pypi.org/project/neurocode-ai/)
+[![Visual Studio Marketplace](https://img.shields.io/visual-studio-marketplace/v/DeepakR.neurocode-ai.svg?label=VS%20Code%20Marketplace)](https://marketplace.visualstudio.com/items?itemName=DeepakR.neurocode-ai)
+[![Glama MCP Server](https://glama.ai/mcp/servers/Arkz-Deepak/neurocode/badge)](https://glama.ai/mcp/servers/Arkz-Deepak/neurocode)
+[![Awesome MCP Servers](https://img.shields.io/badge/Awesome--MCP--Servers-Listed-green.svg)](https://github.com/punkpeye/awesome-mcp-servers)
+[![ICPC 2027](https://img.shields.io/badge/ICPC%202027-Submission%20%2363-purple.svg)](https://conf.researchr.org/home/icpc-2027)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 [![Python 3.10+](https://img.shields.io/badge/python-3.10+-blue.svg)](https://www.python.org/downloads/)
 [![Hardware: 4GB GPU](https://img.shields.io/badge/Hardware-4GB%20VRAM%20Verified-brightgreen.svg)]()
-[![Model Context Protocol](https://img.shields.io/badge/MCP-2024--11--05%20Stdio-blueviolet.svg)](https://modelcontextprotocol.io)
-[![Ollama Integration](https://img.shields.io/badge/Ollama-Offline%20Native-orange.svg)](https://ollama.ai)
-[![Kaggle: Verified](https://img.shields.io/badge/Kaggle-Notebook%20Live-20beff.svg)](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
 
 **NeuroCode** is a lightweight, heterogeneous neuro-symbolic framework and **Model Context Protocol (MCP)** server that enables small, local language models (1.5B–14B parameters) to fix multi-file software engineering bugs on consumer laptops (e.g., 4GB–8GB VRAM) without cloud API dependency or Out-of-Memory (OOM) crashes.
 
@@ -51,11 +36,11 @@ NeuroCode was evaluated across 5 comprehensive benchmark tracks spanning real-wo
 
 | Benchmark Suite | Metric | Baseline / BM25 | NeuroCode (Ours) | Advantage |
 | :--- | :--- | :--- | :--- | :--- |
-| **SWE-bench Lite** ($N=300$) | Resolve Rate (Pass@1) | 0.0% (OOM on 4GB) | **41.4%** (Local) / **21.1%** (14B) | Prevents OOM cliff; +14.2% over BM25 |
-| **HumanEval** ($N=164$) | Pass@1 | 80.5% | **86.0%** (141/164) | +5.5% via syntax-aware graph repair |
-| **MBPP** ($N=100$) | Pass@1 | 81.0% | **87.0%** (87/100) | +6.0% algorithmic verification |
+| **SWE-bench Lite** ($N=292$) | Valid Diff Apply / Resolved | 0.0% (OOM on 4GB) / 0.0% | **41.4% apply / 9.5% resolved** | Prevents OOM cliff; +23.9% apply over BM25 |
+| **HumanEval** ($N=164$) | Pass@1 (Unit Test) | 80.5% | **86.0%** (141/164) | +5.5% via syntax-aware graph repair |
+| **MBPP** ($N=100$) | Pass@1 (Unit Test) | 81.0% | **87.0%** (87/100) | +6.0% algorithmic verification |
 | **RepoBench-R** ($N=150$) | Recall@1 / Token Budget | 18.0% (745 tokens) | **100.0%** (**428 tokens**) | **-42.6% tokens**, perfect cross-file retrieval |
-| **Host Systems Telemetry** | Active VRAM / Slice Latency | 14.2 GB (OOM) | **961.2 MB** / **1.09 ms** | **93.2% VRAM reduction**, 0 OOMs |
+| **Host Systems Telemetry** | Active VRAM / Slice Latency | 14.2 GB (OOM) | **961.2 MB** / **0.62 ms** | **93.2% VRAM reduction**, 0 OOMs |
 
 ### 💻 Physical Laptop Edge Validation (NVIDIA RTX 3050 Laptop GPU)
 Tested live on an everyday developer laptop (AMD Ryzen 7 7435HS + NVIDIA RTX 3050 Laptop GPU, 4.0 GB VRAM):
@@ -124,16 +109,31 @@ Add to your `~/.continue/config.json`:
 
 ---
 
+## 💻 Visual Studio Code Extension
+
+Install directly from the [Visual Studio Marketplace](https://marketplace.visualstudio.com/items?itemName=DeepakR.neurocode-ai):
+
+```bash
+code --install-extension DeepakR.neurocode-ai
+```
+
+*(Or download the `.vsix` binary from the [v0.1.0 Release](https://github.com/Arkz-Deepak/neurocode/releases/tag/v0.1.0))*
+
+### Extension Features:
+- **Status Bar Telemetry**: Live VRAM headroom and `/dev/shm` RAM health indicators.
+- **Context Slicing**: Right-click any active file or code block to extract sub-1,200 token reachability closure in $<100$ ms.
+- **1-Click Prompt Copy**: Instant paste into Cursor, Claude, or Copilot without context overflow.
+- **Autonomous Diff Viewer**: Direct side-by-side patch inspection powered by local Ollama backbones.
+
+---
+
 ## 🚀 Quickstart & CLI Usage
 
-### 1. Installation
+### 1. Installation via PyPI
 ```bash
-git clone https://github.com/Arkz-Deepak/neurocode.git
-cd neurocode
-python3 -m venv .venv
-source .venv/bin/activate
-pip install -e .
+pip install neurocode-ai
 ```
+*(Or install locally from source: `git clone https://github.com/Arkz-Deepak/neurocode.git && cd neurocode && pip install -e .`)*
 
 ### 2. Extract Token-Bounded Slice for Any Issue
 ```bash
@@ -170,10 +170,15 @@ neurocode-copilot
 
 ## 📄 Publications & Preprints
 
-- **IEEE Transactions on Software Engineering (TSE)**: [`paper_ieee_tse/neurocode_tse.tex`](paper_ieee_tse/neurocode_tse.tex)
-- **IEEE Conference Modular Package**: [`paper_modular_ieee/neurocode_modular_paper.tex`](paper_modular_ieee/neurocode_modular_paper.tex)
-- **MLSys Conference Format**: [`paper_mlsys/neurocode_mlsys.tex`](paper_mlsys/neurocode_mlsys.tex)
-- **Kaggle Competition Executable Notebook**: [Kaggle Notebook Live](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
+- **35th IEEE/ACM International Conference on Program Comprehension (ICPC 2027)**:
+  - *Track*: Tool Demonstration Track (Submission #63)
+  - *Title*: *NeuroCode: In-Memory Code Property Graph Slicing and Edge SWE Agent for Resource-Constrained Environments*
+  - *Source*: [`paper_icpc_tool_demo/neurocode_icpc_demo.tex`](paper_icpc_tool_demo/neurocode_icpc_demo.tex)
+- **Journal of Systems and Software (Elsevier)**:
+  - *Track*: Full Research Article (Traditional Subscription, ₹0 APC)
+  - *Title*: *On-Device Autonomous Software Engineering: In-Memory Code Property Graph Reachability Slicing under Edge VRAM Constraints*
+  - *Source*: [`paper_elsevier_jss/neurocode_jss.tex`](paper_elsevier_jss/neurocode_jss.tex)
+- **Kaggle Empirical Reproduction**: [Kaggle Notebook Live](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
 
 ### Citation
 ```bibtex

neurocode_ai-0.1.0/README.md → neurocode_ai-1.2.0/neurocode_ai.egg-info/PKG-INFO

@@ -1,11 +1,32 @@
+Metadata-Version: 2.4
+Name: neurocode-ai
+Version: 1.2.0
+Summary: Heterogeneous Code-Graph Slicing & AST Subtree Grafting for On-Device Autonomous SWE Agents
+Author: Deepak R.
+License: MIT
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: tree-sitter>=0.22.0
+Requires-Dist: tree-sitter-python>=0.21.0
+Requires-Dist: tree-sitter-javascript>=0.21.0
+Requires-Dist: tree-sitter-typescript>=0.21.0
+Requires-Dist: networkx>=3.0
+Requires-Dist: datasets>=2.18.0
+Requires-Dist: matplotlib>=3.8.0
+Requires-Dist: transformers>=4.40.0
+Dynamic: license-file
+
 # NeuroCode: Heterogeneous Code-Graph Slicing for On-Device Autonomous Software Engineering
 
+[![PyPI version](https://img.shields.io/pypi/v/neurocode-ai.svg)](https://pypi.org/project/neurocode-ai/)
+[![Visual Studio Marketplace](https://img.shields.io/visual-studio-marketplace/v/DeepakR.neurocode-ai.svg?label=VS%20Code%20Marketplace)](https://marketplace.visualstudio.com/items?itemName=DeepakR.neurocode-ai)
+[![Glama MCP Server](https://glama.ai/mcp/servers/Arkz-Deepak/neurocode/badge)](https://glama.ai/mcp/servers/Arkz-Deepak/neurocode)
+[![Awesome MCP Servers](https://img.shields.io/badge/Awesome--MCP--Servers-Listed-green.svg)](https://github.com/punkpeye/awesome-mcp-servers)
+[![ICPC 2027](https://img.shields.io/badge/ICPC%202027-Submission%20%2363-purple.svg)](https://conf.researchr.org/home/icpc-2027)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 [![Python 3.10+](https://img.shields.io/badge/python-3.10+-blue.svg)](https://www.python.org/downloads/)
 [![Hardware: 4GB GPU](https://img.shields.io/badge/Hardware-4GB%20VRAM%20Verified-brightgreen.svg)]()
-[![Model Context Protocol](https://img.shields.io/badge/MCP-2024--11--05%20Stdio-blueviolet.svg)](https://modelcontextprotocol.io)
-[![Ollama Integration](https://img.shields.io/badge/Ollama-Offline%20Native-orange.svg)](https://ollama.ai)
-[![Kaggle: Verified](https://img.shields.io/badge/Kaggle-Notebook%20Live-20beff.svg)](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
 
 **NeuroCode** is a lightweight, heterogeneous neuro-symbolic framework and **Model Context Protocol (MCP)** server that enables small, local language models (1.5B–14B parameters) to fix multi-file software engineering bugs on consumer laptops (e.g., 4GB–8GB VRAM) without cloud API dependency or Out-of-Memory (OOM) crashes.
 
@@ -34,11 +55,11 @@ NeuroCode was evaluated across 5 comprehensive benchmark tracks spanning real-wo
 
 | Benchmark Suite | Metric | Baseline / BM25 | NeuroCode (Ours) | Advantage |
 | :--- | :--- | :--- | :--- | :--- |
-| **SWE-bench Lite** ($N=300$) | Resolve Rate (Pass@1) | 0.0% (OOM on 4GB) | **41.4%** (Local) / **21.1%** (14B) | Prevents OOM cliff; +14.2% over BM25 |
-| **HumanEval** ($N=164$) | Pass@1 | 80.5% | **86.0%** (141/164) | +5.5% via syntax-aware graph repair |
-| **MBPP** ($N=100$) | Pass@1 | 81.0% | **87.0%** (87/100) | +6.0% algorithmic verification |
+| **SWE-bench Lite** ($N=292$) | Valid Diff Apply / Resolved | 0.0% (OOM on 4GB) / 0.0% | **41.4% apply / 9.5% resolved** | Prevents OOM cliff; +23.9% apply over BM25 |
+| **HumanEval** ($N=164$) | Pass@1 (Unit Test) | 80.5% | **86.0%** (141/164) | +5.5% via syntax-aware graph repair |
+| **MBPP** ($N=100$) | Pass@1 (Unit Test) | 81.0% | **87.0%** (87/100) | +6.0% algorithmic verification |
 | **RepoBench-R** ($N=150$) | Recall@1 / Token Budget | 18.0% (745 tokens) | **100.0%** (**428 tokens**) | **-42.6% tokens**, perfect cross-file retrieval |
-| **Host Systems Telemetry** | Active VRAM / Slice Latency | 14.2 GB (OOM) | **961.2 MB** / **1.09 ms** | **93.2% VRAM reduction**, 0 OOMs |
+| **Host Systems Telemetry** | Active VRAM / Slice Latency | 14.2 GB (OOM) | **961.2 MB** / **0.62 ms** | **93.2% VRAM reduction**, 0 OOMs |
 
 ### 💻 Physical Laptop Edge Validation (NVIDIA RTX 3050 Laptop GPU)
 Tested live on an everyday developer laptop (AMD Ryzen 7 7435HS + NVIDIA RTX 3050 Laptop GPU, 4.0 GB VRAM):
@@ -107,16 +128,31 @@ Add to your `~/.continue/config.json`:
 
 ---
 
+## 💻 Visual Studio Code Extension
+
+Install directly from the [Visual Studio Marketplace](https://marketplace.visualstudio.com/items?itemName=DeepakR.neurocode-ai):
+
+```bash
+code --install-extension DeepakR.neurocode-ai
+```
+
+*(Or download the `.vsix` binary from the [v0.1.0 Release](https://github.com/Arkz-Deepak/neurocode/releases/tag/v0.1.0))*
+
+### Extension Features:
+- **Status Bar Telemetry**: Live VRAM headroom and `/dev/shm` RAM health indicators.
+- **Context Slicing**: Right-click any active file or code block to extract sub-1,200 token reachability closure in $<100$ ms.
+- **1-Click Prompt Copy**: Instant paste into Cursor, Claude, or Copilot without context overflow.
+- **Autonomous Diff Viewer**: Direct side-by-side patch inspection powered by local Ollama backbones.
+
+---
+
 ## 🚀 Quickstart & CLI Usage
 
-### 1. Installation
+### 1. Installation via PyPI
 ```bash
-git clone https://github.com/Arkz-Deepak/neurocode.git
-cd neurocode
-python3 -m venv .venv
-source .venv/bin/activate
-pip install -e .
+pip install neurocode-ai
 ```
+*(Or install locally from source: `git clone https://github.com/Arkz-Deepak/neurocode.git && cd neurocode && pip install -e .`)*
 
 ### 2. Extract Token-Bounded Slice for Any Issue
 ```bash
@@ -153,10 +189,15 @@ neurocode-copilot
 
 ## 📄 Publications & Preprints
 
-- **IEEE Transactions on Software Engineering (TSE)**: [`paper_ieee_tse/neurocode_tse.tex`](paper_ieee_tse/neurocode_tse.tex)
-- **IEEE Conference Modular Package**: [`paper_modular_ieee/neurocode_modular_paper.tex`](paper_modular_ieee/neurocode_modular_paper.tex)
-- **MLSys Conference Format**: [`paper_mlsys/neurocode_mlsys.tex`](paper_mlsys/neurocode_mlsys.tex)
-- **Kaggle Competition Executable Notebook**: [Kaggle Notebook Live](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
+- **35th IEEE/ACM International Conference on Program Comprehension (ICPC 2027)**:
+  - *Track*: Tool Demonstration Track (Submission #63)
+  - *Title*: *NeuroCode: In-Memory Code Property Graph Slicing and Edge SWE Agent for Resource-Constrained Environments*
+  - *Source*: [`paper_icpc_tool_demo/neurocode_icpc_demo.tex`](paper_icpc_tool_demo/neurocode_icpc_demo.tex)
+- **Journal of Systems and Software (Elsevier)**:
+  - *Track*: Full Research Article (Traditional Subscription, ₹0 APC)
+  - *Title*: *On-Device Autonomous Software Engineering: In-Memory Code Property Graph Reachability Slicing under Edge VRAM Constraints*
+  - *Source*: [`paper_elsevier_jss/neurocode_jss.tex`](paper_elsevier_jss/neurocode_jss.tex)
+- **Kaggle Empirical Reproduction**: [Kaggle Notebook Live](https://www.kaggle.com/code/deepark121/neurocode-heterogeneous-graph-sliced-swe-agent)
 
 ### Citation
 ```bibtex

{neurocode_ai-0.1.0 → neurocode_ai-1.2.0}/neurocode_ai.egg-info/SOURCES.txt

@@ -10,6 +10,7 @@ neurocode_ai.egg-info/top_level.txt
 src/__init__.py
 src/cli.py
 src/mcp_server.py
+src/agent/graft.py
 src/agent/llm_interface.py
 src/agent/patcher.py
 src/benchmark/create_thumbnail.py
@@ -31,5 +32,8 @@ src/graph/__init__.py
 src/graph/builder.py
 src/graph/parser.py
 src/graph/slicer.py
+tests/test_graft.py
 tests/test_graph.py
-tests/test_multi_hop.py
+tests/test_mcp.py
+tests/test_multi_hop.py
+tests/test_ts_js.py

{neurocode_ai-0.1.0 → neurocode_ai-1.2.0}/neurocode_ai.egg-info/requires.txt

@@ -1,5 +1,7 @@
 tree-sitter>=0.22.0
 tree-sitter-python>=0.21.0
+tree-sitter-javascript>=0.21.0
+tree-sitter-typescript>=0.21.0
 networkx>=3.0
 datasets>=2.18.0
 matplotlib>=3.8.0

{neurocode_ai-0.1.0 → neurocode_ai-1.2.0}/pyproject.toml

@@ -4,8 +4,8 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "neurocode-ai"
-version = "0.1.0"
-description = "Heterogeneous Code-Graph Slicing for On-Device Autonomous SWE Agents"
+version = "1.2.0"
+description = "Heterogeneous Code-Graph Slicing & AST Subtree Grafting for On-Device Autonomous SWE Agents"
 readme = "README.md"
 authors = [{ name = "Deepak R." }]
 license = { text = "MIT" }
@@ -13,6 +13,8 @@ requires-python = ">=3.10"
 dependencies = [
     "tree-sitter>=0.22.0",
     "tree-sitter-python>=0.21.0",
+    "tree-sitter-javascript>=0.21.0",
+    "tree-sitter-typescript>=0.21.0",
     "networkx>=3.0",
     "datasets>=2.18.0",
     "matplotlib>=3.8.0",

neurocode_ai-1.2.0/src/agent/graft.py

@@ -0,0 +1,241 @@
+"""
+AST Subtree Grafting Engine for Autonomous SWE Agents.
+Performs deterministic, syntax-verified AST node replacements in-memory,
+completely eliminating fragile multi-line whitespace and indentation diff failures.
+"""
+
+import ast
+import difflib
+import os
+from pathlib import Path
+from typing import Dict, Any, Optional, Tuple
+from tree_sitter import Node, Parser, Language
+import tree_sitter_python as tspython
+
+_PY_LANG = Language(tspython.language())
+_PY_PARSER = Parser(_PY_LANG)
+
+try:
+    import tree_sitter_javascript as tsjs
+    _JS_LANG = Language(tsjs.language())
+    _JS_PARSER = Parser(_JS_LANG)
+except Exception:
+    _JS_LANG = None
+    _JS_PARSER = None
+
+try:
+    import tree_sitter_typescript as tsts
+    _TS_LANG = Language(tsts.language_typescript())
+    _TS_PARSER = Parser(_TS_LANG)
+    _TSX_LANG = Language(tsts.language_tsx())
+    _TSX_PARSER = Parser(_TSX_LANG)
+except Exception:
+    _TS_LANG = None
+    _TS_PARSER = None
+    _TSX_LANG = None
+    _TSX_PARSER = None
+
+
+def get_parser_for_file(file_path: str) -> Parser:
+    """Returns the appropriate Tree-sitter parser based on file extension."""
+    ext = Path(file_path).suffix.lower()
+    if ext == ".tsx" and _TSX_PARSER:
+        return _TSX_PARSER
+    elif ext == ".ts" and _TS_PARSER:
+        return _TS_PARSER
+    elif ext in (".js", ".jsx", ".mjs", ".cjs") and _JS_PARSER:
+        return _JS_PARSER
+    return _PY_PARSER
+
+
+def find_target_node(root: Node, symbol: str, code_bytes: bytes) -> Optional[Node]:
+    """
+    Locates the AST node corresponding to the given symbol name or 'ClassName.methodName'.
+    """
+    def get_text(node: Optional[Node]) -> str:
+        if not node:
+            return ""
+        return code_bytes[node.start_byte:node.end_byte].decode("utf-8", errors="replace")
+
+    target_class = None
+    target_name = symbol
+    if "." in symbol:
+        target_class, target_name = symbol.split(".", 1)
+
+    stack = [root]
+    while stack:
+        curr = stack.pop()
+        
+        # Check Python Function
+        if curr.type in ("function_definition", "function_declaration", "method_definition"):
+            name_node = curr.child_by_field_name("name")
+            if name_node and get_text(name_node) == target_name:
+                if target_class:
+                    # Verify parent class
+                    parent = curr.parent
+                    while parent and parent.type not in ("class_definition", "class_declaration"):
+                        parent = parent.parent
+                    if parent:
+                        p_name_node = parent.child_by_field_name("name")
+                        if p_name_node and get_text(p_name_node) == target_class:
+                            return curr
+                else:
+                    return curr
+
+        # Check Python or JS/TS Class
+        elif curr.type in ("class_definition", "class_declaration"):
+            name_node = curr.child_by_field_name("name")
+            if not target_class and name_node and get_text(name_node) == target_name:
+                return curr
+
+        # Check JS/TS Arrow Function or Variable
+        elif curr.type == "variable_declarator":
+            name_node = curr.child_by_field_name("name")
+            val_node = curr.child_by_field_name("value")
+            if name_node and get_text(name_node) == target_name:
+                # Return the declarator's parent lexical_declaration if possible
+                if curr.parent and curr.parent.type in ("lexical_declaration", "variable_declaration"):
+                    return curr.parent
+                return curr
+
+        for child in reversed(curr.children):
+            stack.append(child)
+
+    return None
+
+
+def align_indentation(replacement_code: str, target_indent_spaces: int) -> str:
+    """
+    Normalizes and re-indents replacement code so it seamlessly matches the AST target scope.
+    """
+    lines = replacement_code.strip("\r\n").splitlines()
+    if not lines:
+        return ""
+
+    # Determine common leading whitespace in replacement lines (excluding empty lines)
+    non_empty = [l for l in lines if l.strip()]
+    if not non_empty:
+        return replacement_code
+
+    min_leading = min(len(l) - len(l.lstrip()) for l in non_empty)
+    unindented_lines = [(l[min_leading:] if len(l) >= min_leading else l) for l in lines]
+
+    target_indent = " " * target_indent_spaces
+    reindented = []
+    for idx, line in enumerate(unindented_lines):
+        if not line.strip():
+            reindented.append("")
+        elif idx == 0:
+            reindented.append(line.lstrip())
+        else:
+            leading_len = len(line) - len(line.lstrip())
+            reindented.append(target_indent + (" " * leading_len) + line.lstrip())
+
+    return "\n".join(reindented)
+
+
+def graft_symbol(
+    repo_path: str,
+    file_path: str,
+    symbol: str,
+    replacement_code: str,
+    apply: bool = False
+) -> Dict[str, Any]:
+    """
+    Performs deterministic AST subtree grafting for a target symbol.
+    """
+    base_dir = Path(repo_path).resolve()
+    target_file = (base_dir / file_path).resolve()
+
+    if not target_file.exists():
+        return {
+            "success": False,
+            "error": f"File does not exist: {file_path}",
+            "file_path": file_path
+        }
+
+    try:
+        original_source = target_file.read_text(encoding="utf-8", errors="replace")
+    except Exception as e:
+        return {"success": False, "error": f"Failed to read file: {e}", "file_path": file_path}
+
+    parser = get_parser_for_file(str(target_file))
+    code_bytes = original_source.encode("utf-8")
+    tree = parser.parse(code_bytes)
+
+    # 1. Locate Target AST Node
+    target_node = find_target_node(tree.root_node, symbol, code_bytes)
+    if not target_node:
+        return {
+            "success": False,
+            "error": f"Symbol '{symbol}' was not found in the AST of {file_path}.",
+            "file_path": file_path
+        }
+
+    # 2. Indentation alignment
+    target_indent_cols = target_node.start_point[1]
+    formatted_replacement = align_indentation(replacement_code, target_indent_cols)
+
+    # 3. Perform byte slice replacement
+    new_bytes = (
+        code_bytes[:target_node.start_byte]
+        + formatted_replacement.encode("utf-8")
+        + code_bytes[target_node.end_byte:]
+    )
+    new_source = new_bytes.decode("utf-8", errors="replace")
+
+    # 4. AST Integrity Verification
+    new_tree = parser.parse(new_bytes)
+    if new_tree.root_node.has_error:
+        return {
+            "success": False,
+            "error": "SyntaxError: Grafted code resulted in invalid AST grammar.",
+            "file_path": file_path,
+            "syntax_valid": False
+        }
+
+    # Python-specific AST check
+    if target_file.suffix == ".py":
+        try:
+            ast.parse(new_source, filename=str(target_file))
+        except SyntaxError as se:
+            return {
+                "success": False,
+                "error": f"Python SyntaxError at line {se.lineno}: {se.msg}",
+                "file_path": file_path,
+                "syntax_valid": False
+            }
+
+    # 5. Generate Diff
+    diff_lines = list(difflib.unified_diff(
+        original_source.splitlines(keepends=True),
+        new_source.splitlines(keepends=True),
+        fromfile=f"a/{file_path}",
+        tofile=f"b/{file_path}",
+        lineterm=""
+    ))
+    diff_str = "\n".join(diff_lines)
+
+    # 6. Apply to disk if requested
+    if apply:
+        try:
+            target_file.write_text(new_source, encoding="utf-8")
+        except Exception as e:
+            return {
+                "success": False,
+                "error": f"Failed to write changes to disk: {e}",
+                "file_path": file_path
+            }
+
+    return {
+        "success": True,
+        "file_path": file_path,
+        "symbol": symbol,
+        "entity_type": target_node.type,
+        "start_line": target_node.start_point[0] + 1,
+        "end_line": target_node.end_point[0] + 1,
+        "syntax_valid": True,
+        "applied": apply,
+        "diff": diff_str,
+        "message": "Subtree successfully grafted with 100% verified syntax."
+    }