Skip to content

Security: intutic/turing

.github/SECURITY.md

Security Policy & Vulnerability Disclosure

Turing Engine and Intutic take the security and integrity of our codebase, native SIMD binaries, and cryptographic model containers seriously.


πŸ›‘οΈ Supported Versions

We provide active security patches, bug fixes, and vulnerability remediation for the following release series:

Version Series Release Date Supported Status
v0.2.x / v0.1.x August 2026 – Present 🟒 Active Security Support
< v0.1.0 Experimental πŸ”΄ End of Life

πŸ”’ Reporting a Vulnerability

If you discover a security vulnerability, side-channel attack, memory safety defect in our C++20 AVX2 native extensions (csrc/), or an authentication/authorization bypass in the serving endpoints:

1. Private Vulnerability Reporting (Preferred)

Please submit a report directly via GitHub Private Vulnerability Reporting. This keeps the disclosure private between you and our security maintainers until a fix is released.

2. Direct Security Contact

Alternatively, you can email our security response team directly:

  • Email: security@intutic.ai / support@intutic.ai
  • Lead Maintainer: Ishan Gupta (ishangupta.ds@gmail.com)

πŸ“‹ What to Include in Your Report

To help us triage and resolve the issue quickly, please provide:

  1. Description: A clear overview of the potential vulnerability.
  2. Impact: Which components are affected (e.g. csrc/, /v1/chat/completions, license_gate.py, .tgate binary deserializer).
  3. Reproduction Steps / PoC: Minimal reproducible Python/C++ script or payload demonstrating the issue.
  4. Environment: OS (Linux, macOS, Windows), Python version, and CUDA/AVX2 hardware configuration.

⏱️ Response Timelines & Policy

  • Initial Response: Within 24 hours acknowledging receipt of your report.
  • Triage & Assessment: Within 48 hours with a severity rating (CVSS score) and reproduction confirmation.
  • Remediation & Patch: Target fix released within 7 to 14 business days depending on severity.
  • Public Disclosure: Coordinated disclosure after the patch has been published and validated.

🚫 Non-Eligible Issues

  • Social engineering, phishing, or physical attacks.
  • Denial of Service (DoS) via resource exhaustion on unauthenticated local debug endpoints without production gateway deployment.
  • Issues in third-party upstream dependencies (please report directly to the respective upstream project, e.g., PyTorch, FastAPI, Triton).

There aren't any published security advisories