# 🦙 Run KIN Cyber v2 with Ollama (Local & Offline)

Security engineers and penetration testers can run **KIN Cyber v2** 100% locally and offline without leaking sensitive target code to third-party APIs.

---

## ⚡ Option 1: One-Line Run via Hugging Face Integration

```bash
# Pull and run instantly in terminal
ollama run hf.co/nyxspecter4/kin-sft-lora
```

---

## 🛠️ Option 2: Build Custom Local Agent via Modelfile

1. Create a file named `Modelfile` with the following contents:

```dockerfile
FROM Qwen/Qwen2.5-3B-Instruct
ADAPTER nyxspecter4/kin-sft-lora

TEMPLATE """{{ if .System }}<|im_start|>system
{{ .System }}<|im_end|>
{{ end }}{{ if .Prompt }}<|im_start|>user
{{ .Prompt }}<|im_end|>
{{ end }}<|im_start|>assistant
"""

SYSTEM """You are KIN — an elite agentic cybersecurity partner. Direct, opinionated, and specific. Name tools, CVEs, companies, and invariants."""
PARAMETER temperature 0.3
PARAMETER top_p 0.9
```

2. Build and launch:
```bash
ollama create kin-cyber -f Modelfile
ollama run kin-cyber
```

---

## 📦 Quantization Matrix

| Quant Format | Precision | File Size | Recommended Hardware |
|---|---|---|---|
| **Q4_K_M** | 4-bit Medium | ~2.1 GB | Laptops & M1/M2/M3 Macs (Fastest) |
| **Q5_K_M** | 5-bit Medium | ~2.5 GB | Standard Desktops (Balanced) |
| **Q8_0** | 8-bit High | ~3.8 GB | Workstations / Servers (Maximum Fidelity) |
| **FP16** | 16-bit Full | ~6.2 GB | GPU VRAM >= 8 GB (Uncompressed) |