add README.md

README.md

@@ -1,165 +1,13 @@
 ---
+library_name: mlx
 license: apache-2.0
+license_link: https://huggingface.co/Qwen/Qwen3-8B/blob/main/LICENSE
+pipeline_tag: text-generation
 base_model: georgehenney/Qwen3-8B-heretic
 tags:
-  - security
-  - cybersecurity
-  - pentest
-  - vulnerability-assessment
-  - CVSS
-  - NIST
-  - OWASP
-  - red-team
-  - bug-bounty
-  - 128k-context
-  - yarn
-  - MLX
-  - Safetensors
-  - qwen3
-  - heretic
-  - 4-bit
-  - precision
-  - quantized
-  - apple-silicon
-  - reasoning
-  - chain-of-thought
-  - ravenx
-  - tool-calling
-  - function-calling
-  - rath-protocol
-  - conversational
-language:
-  - en
-pipeline_tag: text-generation
-library_name: mlx
----
-
-# RavenX-Sec 8B — Security RATH Protocol 128K (MLX 4-bit)
-
-<p align="center">
-<strong>🔒 Find → Classify → Fix → Verify → Report</strong><br>
-<em>Autonomous Security Intelligence Model. 128K context. 379K training examples. 15 security datasets. RATH protocol.</em>
-</p>
-
-**Built by [@DeadByDawn101](https://github.com/DeadByDawn101) (RavenX LLC)**
-
-Part of the [RavenX MLX Models — Apple Silicon Inference Stack](https://huggingface.co/collections/deadbydawn101/ravenx-mlx-models-apple-silicon-inference-stack-67f7270ac5b85c49a6c0f47f) collection.
-
-## Quick Start
-
-```python
-from mlx_lm import load, generate
-
-model, tokenizer = load("deadbydawn101/RavenX-Sec-8B-Security-RATH-128k-mlx-4bit")
-
-messages = [
-    {"role": "system", "content": "You are RavenX-Sec. Follow the RATH protocol for every finding."},
-    {"role": "user", "content": "You found OpenSSH 7.4 on port 22. Classify and remediate."}
-]
-
-prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
-response = generate(model, tokenizer, prompt=prompt, max_tokens=1024)
-print(response)
-```
-
-### Chat REPL
-```bash
-mlx_lm.chat --model "deadbydawn101/RavenX-Sec-8B-Security-RATH-128k-mlx-4bit"
-```
-
-### OpenAI-Compatible Server
-```bash
-mlx_lm.server --model "deadbydawn101/RavenX-Sec-8B-Security-RATH-128k-mlx-4bit"
-```
-
-### With Hermes Agent
-```bash
-hermes config set model.base_url http://127.0.0.1:8080/v1
-hermes config set model.default deadbydawn101/RavenX-Sec-8B-Security-RATH-128k-mlx-4bit
-```
-
-## What This Is
-
-RavenX-Sec is a fine-tuned security model that produces structured vulnerability assessments using the **RATH protocol** — citing specific CVEs, generating CVSS vectors, showing vulnerable code, providing exploit PoCs, and delivering actionable remediation.
-
-Trained on **379,286 examples from 15 security datasets** with **84% security-focused content**. Extended to **128K context** via YaRN rope scaling for full pentest engagement support.
-
-## RATH Protocol Output
-
-```
-RATH STEP 1: IDENTIFY
-- CVE: CVE-2016-10009, CVE-2016-10010, CVE-2016-10011
-- Description: OpenSSH 7.4 — authentication bypass via SSH agent forwarding
-
-RATH STEP 2: ASSESS
-- CVSS Score: 6.3
-- CVSS Vector: CVSS:3.0/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:L
-- Technical Analysis: SSH agent forwarding vulnerabilities
-
-RATH STEP 3: THREAT
-- Vulnerable Code: ForwardAgent yes (VULNERABLE) → ForwardAgent no (SAFE)
-- Exploit PoC: ssh -o ForwardAgent=yes user@host
-
-RATH STEP 4: REMEDIATE
-- Disable SSH agent forwarding unless necessary
-- Restrict to specific users via Match blocks
-- Commands: sudo apt upgrade openssh-server -y
-```
-
-## Model Details
-
-| Parameter | Value |
-|-----------|-------|
-| **Base** | georgehenney/Qwen3-8B-heretic (abliterated Qwen3-8B) |
-| **Quantization** | MLX 4-bit |
-| **Context Window** | 128K (YaRN rope scaling, factor 4.0) |
-| **Training Data** | 379,286 examples (84% security) |
-| **Datasets** | 15 security-specific |
-| **Tokens Trained** | 3,019,210 |
-| **Method** | MLX LoRA (rank 32, 8 layers, 1e-5 LR) |
-| **Hardware** | Apple M4 Max 128GB |
-| **Iterations** | 1500 |
-
-## Training Datasets (15)
-
-**Offensive / Red Team:** WNT3D/Ultimate-Offensive-Red-Team (25.6K) • Rootkit7/pentest-redteam-steering (2K) • acnimatic3722/kali-linux-pentesting-data (343)
-
-**Pentest / Bug Bounty:** SkywardNomad92/pentest-findings-v2 (50K) • theelderemo/pentesting-explanations (5.9K) • AYI-NEDJIMI/bug-bounty-pentest-en (27) • CJJones/Synthetic_PenTest_Reports (123)
-
-**CVE / Vulnerability:** auren-research/cve-sft-v5 (10K) • Trendyol/Cybersecurity-Instruction-Tuning (50K)
-
-**Agentic / Coding:** WithinUsAI/AgentAngel_100k (50K capped) • WithinUsAI/claude_mythos_distilled_25k (16K security)
-
-**Extracted:** hackingBuddyGPT • PentestGPT • Shannon • Ghidra + Synthetic RATH chains
-
-## Frameworks Supported
-
-CVSS 3.1 • NIST CSF 2.0 • OWASP Top 10 • CWE • MITRE ATT&CK • PCI DSS • HIPAA • SOX
-
-## Use Cases
-
-- **Authorized Penetration Testing** — Full engagement with RATH-classified findings
-- **Bug Bounty Hunting** — Automated vulnerability analysis and report generation
-- **CVE Triage** — Classify and prioritize vulnerability scan results
-- **Hardening Review** — Analyze configs against CIS Benchmarks
-- **Incident Response** — Containment, eradication, and recovery guidance
-- **Security Training** — Learn offensive and defensive techniques
-
-## Related Models
-
-| Model | Format | Link |
-|-------|--------|------|
-| **RavenX-Sec 8B MLX 4-bit** | MLX Safetensors | This repo |
-| RavenX-Sec 8B GGUF | GGUF (F16/Q8/Q5/Q4) | [deadbydawn101/RavenX-Sec-8B-GGUF](https://huggingface.co/deadbydawn101/RavenX-Sec-8B-GGUF) |
-
-## Source Code
-
-[github.com/DeadByDawn101/RavenX-Sec](https://github.com/DeadByDawn101/RavenX-Sec)
-
-## License
-
-Apache-2.0
-
+- heretic
+- uncensored
+- decensored
+- abliterated
+- mlx
 ---
-
-*"We don't give up. We do what others don't and build what isn't possible." — RavenX LLC*