NeuroGraph

High-performance cognitive platform with Rust Core, WebSocket API, and Jupyter integration

What is NeuroGraph?

NeuroGraph is a cognitive computing platform that combines:

• Rust Core - High-performance event processing (304K events/sec, 0.39μs latency)
• WebSocket API - Real-time bidirectional communication (~5ms latency)
• Jupyter Integration - Interactive notebooks with magic commands
• Web Dashboard - React SPA with real-time monitoring

Quick Start

Installation

From PyPI:

pip install ngcore              # Core package
pip install ngcore[jupyter]     # With Jupyter integration
pip install ngcore[api]         # With WebSocket API
pip install ngcore[all]         # Full installation

From Source:

# Clone repository
git clone https://github.com/dchrnv/neurograph.git
cd neurograph

# Install dependencies
pip install -e ".[all]"  # Full installation

Usage

Jupyter Notebook:

%load_ext neurograph_jupyter
%neurograph init --path ./my_graph.db
%neurograph query "find all nodes"

Python API:

from neurograph import NeuroGraph

# Your code here

WebSocket Client:

from neurograph_client import WebSocketClient

client = WebSocketClient("ws://localhost:8000/ws")
await client.subscribe("metrics")

Features

• ✅ Rust Core - 304K events/sec processing
• ✅ WebSocket API - Real-time events with ~5ms latency
• ✅ Jupyter Integration - Magic commands and widgets
• ✅ Web Dashboard - React SPA with monitoring
• ✅ Module Registry - Dynamic module management
• ✅ RBAC - Role-based access control
• ✅ CI/CD - GitHub Actions with pytest and cargo test

Documentation

• Documentation: docs/
• Roadmap: ROADMAP.md
• Deferred Tasks: DEFERRED.md
• Changelog: CHANGELOG.md
• Current Status: STATUS.md

Development

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest tests/

# Run Rust tests
cd src/core_rust && cargo test

# Build package
maturin build --release

Project Status

Current: v1.0.0 - Production Ready! 🎉

Recent Updates (2026-01-14):

• ✅ Fixed all Rust compilation issues (0 warnings)
• ✅ Feedback system P1 (rewards) - working
• ✅ Critical discovery: Graph is dead code (1400+ LOC unused)
• ✅ Documented P2 requirements for v1.1.0
• ✅ Fixed documentation broken links
• ✅ Created comprehensive analysis docs (GRAPH_ANALYSIS.md)

Completed Phases:

• ✅ Phase 1: Code Quality (100%)
• ✅ Phase 2: Documentation (100%)
• ✅ Phase 3: Production Readiness (100%)
• ✅ Phase 4: Final Polish (100%)
  • ✅ Phase 4.1: Code Cleanup
  • ✅ Phase 4.2: Documentation Review
  • ✅ Phase 4.3: Release Preparation

Key Features:

• 304K events/sec processing (Rust Core)
• ~5ms WebSocket latency
• 378 comprehensive API tests
• 0 critical security vulnerabilities
• 100M tokens stress tested
• Production monitoring ready (Grafana + Prometheus)
• Docker images <300MB

Known Limitations (v1.0.0):

• ✅ Feedback P1 (reward updates) - working (updates last 1000 events)
• Feedback P2 (user connections) - stub, full implementation v1.1.0
• ADNA proposal application - deferred to v1.1.0
• Graph code (1400+ LOC) - unused, will be removed in v1.1.0
• See DEFERRED.md for full list and GRAPH_ANALYSIS.md for details

Roadmap:

• ✅ v1.0.0: Production Ready (Released!)
• ⏳ v1.1.0: Feedback P2, Graph removal, ADNA proposals (2-3 weeks)
• ⏳ v1.2.0: ConnectionV3 typed links, NN search improvements (2-3 weeks)

See ROADMAP.md and CHANGELOG.md for details.

Contributing

Contributions are welcome! Please read our contributing guidelines before submitting PRs.

License

AGPL-3.0 - See LICENSE file for details.

Links

• Repository: https://github.com/dchrnv/neurograph
• Documentation: docs/
• Issues: https://github.com/dchrnv/neurograph/issues