neuralspace-ai 4.0.0

Cognitive Code Security Engine - Self-evolving, AI-powered threat detection

🧠 NeuralSpace – Cognitive Security Universe

NeuralSpace is the world's first self-organizing, zero-trust security universe for code. It doesn't just scan for known threats—it builds a living, evolving topology of your codebase where every branch has its own specialized "neural brain."

It combines a Covalent Tree (self-evolving topology), a Hive Mind (emergent intelligence), a Zero-Trust Mesh (cryptographic trust), and AST/CFG Taint Analysis (real data-flow tracking) into a single, ultra-lightweight (~8 KB) system.

---

🔥 The Problem We Solve

Current Tool	Limitation	NeuralSpace Advantage
Traditional AV	Relies on known signatures.	Blocks zero‑day obfuscated threats.
SAST (SonarQube)	99.5% false positives.	Contextual detection + Taint Analysis (e.g., requests.get alone is safe; requests.get + exec is a threat).
Transformer Models	Huge, slow, cloud‑dependent.	Lightweight (~8 KB), runs instantly on CPU.
File Watchers	React to files, don't understand content.	Routes files dynamically into a living knowledge tree (the Covalent Tree).

---

✨ Key Features

• 🧬 Self‑Evolving Topology (The Covalent Tree) – The tree spawns new branches anticipatorily when it detects structural drift (drift velocity > 0.5). It doesn't just classify code; it organizes your codebase into a living taxonomy.
• 🧠 Distributed Neural Atoms – Each tree branch has its own PureNeuralAtom (512→4→4 network) initialized with a unique seed. This creates specialized "brains" for different code families.
• 🤝 Hive Mind (Emergent Intelligence) – Multiple agents communicate and form a consensus on threats. The collective intelligence (consensus ≥ 0.7) can override individual node decisions.
• 🛡️ Zero-Trust Security Mesh – All threat reports are cryptographically signed with RSA. Nodes earn trust over time; low-trust nodes (score < 0.3) are ignored.
• 🔍 AST/CFG Taint Analysis – Tracks whether tainted data (user input, network data) reaches dangerous sinks (exec, eval, os.system). Real data-flow analysis, not just token matching.
• 🌍 Polyglot – Scans Python, JavaScript, TypeScript, Go, and Rust (with Tree-Sitter AST parsing).
• ⚡ Ultra‑Lightweight & Local – Trains in under 60 seconds on a standard CPU. No cloud. No GPU. (~8 KB model).
• 🤖 GitHub App Integration – Auto‑scans Pull Requests and posts comments with detailed decision traces.
• 🌐 Federated Intelligence – Global aggregator shares anonymized threat signatures across instances, creating a living immune system.
• 🗣️ God User Interface – Natural language commands to shape the universe (health, spawn branch, show threats, evolve).

---

🏗️ How It Works

1. Tokenization + Taint Analysis – Code is parsed via Tree-Sitter AST, and data-flow taint analysis tracks user input to dangerous sinks.
2. Routing – The vector descends the Covalent Tree. If it matches a child node (cosine similarity > 0.85), it dives deeper. Otherwise, it stops.
3. Hive Mind Consensus – All active nodes vote on the threat. The collective decision overrides individual errors.
4. Judgment – The terminal node's PureNeuralAtom computes two scores:
   • Sentinel (S) – Threat probability (class 3).
   • Logic (L) – Safe probability (class 0).
5. Enforcement – If S > 0.25 or L < 0.2, the file is quarantined and cryptographically reported.
6. Evolution – If the file is allowed but deviates significantly (drift velocity > 0.5), the tree anticipatorily fractures and spawns a new child node.

---

🚀 Quick Start (Global Install)

Installation (One Command)

pip install neuralspace-ai

## Basic Usage

```bash

###  Scan a project folder
neuralspace scan ./your_project --quarantine rename

### Watch a folder in real-time
neuralspace watch ./your_project

###  Sync with the global threat intelligence network
neuralspace sync

### Advanced: Train the AI (Optional)
### The package comes pre-trained. But you can retrain it on your own dataset:

```bash
neuralspace generate
neuralspace train