feat: add SECURITY.md (#1)

Author: abhiprd2000

SECURITY.md

@@ -0,0 +1,54 @@
+# Security Policy
+
+## Overview
+Axiom OS is a research prototype focused on proportional-cost per-read provenance and file integrity. While security is a first-class citizen in the design of the provenance layer, the kernel is currently in an **alpha/research state**. 
+
+We welcome security research, vulnerability reports, and feedback that help strengthen the architecture.
+
+## Supported Versions
+
+Currently, security updates and patches are only provided for the latest development branch.
+
+| Version | Supported          |
+| ------- | ------------------ |
+| v0.3.0  | :white_check_mark: |
+| < v0.3.0| :x:                |
+
+## Reporting a Vulnerability
+
+**Please do not report security vulnerabilities through public GitHub issues.** 
+
+If you discover a security vulnerability, please report it through one of the following methods:
+
+1. **Email:** 8292aniarc@gmail.com
+2. **GitHub Private Vulnerability Reporting:** Please use the "Report a vulnerability" button under the **Security and Quality** tab.
+
+### What to Include
+To help us triage the report quickly, please include:
+- A description of the vulnerability and its potential impact.
+- Steps to reproduce the issue (ideally within the QEMU environment).
+- Any proof-of-concept (PoC) code or shell commands (e.g., using the `tamper` command to show a bypass).
+
+## Our Disclosure Process
+Once a report is received:
+1. We will acknowledge receipt of your report within 48-72 hours.
+2. We will investigate the issue and confirm the vulnerability.
+3. If confirmed, we will work on a fix in a private branch.
+4. We will coordinate a disclosure date with you before making the fix public.
+
+## Threat Model & Scope
+Please refer to the **Threat Model** section in the `README.md` for the current security assumptions. 
+
+### In-Scope for Research:
+- Bypassing the BLAKE3 block-level integrity verification.
+- Escalating privileges from the Shell/User mode to Kernel mode.
+- Breaking the CR3-based process isolation.
+- Vulnerabilities in the Mitra DSL execution path.
+
+### Out-of-Scope:
+- Hardware-level attacks (e.g., Rowhammer) that are explicitly excluded from the current threat model.
+- Denial of Service (DoS) attacks that simply crash the kernel (since it is a single-core research prototype).
+- Issues requiring physical access or DMA-capable peripherals.
+
+## Acknowledgments
+We value the work of independent security researchers. If you report a valid vulnerability that leads to a fix, we are happy to credit you.