Skip to content

Security: RadTome/QuickSnipe-support

Security

SECURITY.md

πŸ”’ Security & Vulnerability Disclosure Policy

Security and data privacy are core architectural foundations of QuickSnipe. Because QuickSnipe is engineered as a 100% client-side Chrome extension with zero remote user databases, we prioritize local sandboxing, strict Content Security Policies (CSP), and secure client-to-API communication.


πŸ›‘οΈ Supported Versions

We actively provide security patches and updates for the following versions:

Version Supported Notes
1.5.0 βœ… Current Live Production Release (Chrome Web Store)
1.4.5 βœ… Previous Stable Release
1.4.4 βœ… Previous Stable Release
1.4.3 βœ… Previous Stable Release
1.4.2 βœ… Previous Stable Release
1.4.1 βœ… Supported Stable Release
1.4.0 βœ… Supported Stable Release
1.3.x ⚠️ Previous Release. Please update to latest version.
< 1.3.0 ❌ Deprecated. Unsupported.

πŸ” Architectural Security Highlights

  1. Least-Privilege Opt-In Footprint: Zero broad wildcard permissions (http://*/* and https://*/* removed). In v1.5.0+, extensionpay.com is the sole required host permission at install; marketplaces (including admin.shopify.com) and cloud AI API endpoints are optional_host_permissions granted strictly per platform/provider on opt-in and revocable anytime.
  2. Zero Intermediate Servers: QuickSnipe does not maintain backend proxies or logging servers. Your search terms, listing copy, and scraped data stay strictly within your local machine.
  3. Encrypted Local Storage & Sandboxing: API keys, swipe files, and QuickBar preferences are stored locally in Chrome's sandboxed chrome.storage.local with serialized writes to prevent race conditions. Zero host-origin storage footprint.
  4. Prompt-Injection Delimiter Hardening: Scraped marketplace fields are wrapped in <untrusted_marketplace_data> delimiters with breakout-tag neutralization before reaching AI prompts.
  5. Bundled Local Fonts & Strict Content Security Policy: QuickSnipe adheres to Manifest V3 CSP (no eval(), no new Function(), zero remote script or font CDN calls β€” all fonts ship inside assets/fonts/).
  6. CSV Formula-Injection Neutralization: Exported CSV spreadsheets escape leading formula characters (=, +, -, @) and emit a UTF-8 BOM.
  7. Direct HTTPS Communication: Optional BYOK API calls connect directly from your client browser to official provider endpoints (generativelanguage.googleapis.com, api.openai.com, api.groq.com, api.anthropic.com, api.deepseek.com, openrouter.ai) only when you configure and authorize that provider.

🚨 Reporting a Security Vulnerability

If you discover a security vulnerability or sensitive data leakage issue within QuickSnipe, please report it responsibly so we can resolve it before public disclosure.

Reporting Procedure:

  1. Do NOT open a public GitHub issue for undisclosed security vulnerabilities.
  2. Submit a report through GitHub Private Vulnerability Reporting: πŸ‘‰ Report a Vulnerability (Private Advisory) (Or navigate to the Security tab at the top of this repository and click "Report a vulnerability").
  3. Include the following details in the private advisory:
    • Type of vulnerability (e.g., XSS, storage exposure, permission escalation).
    • Step-by-step reproduction steps or Proof of Concept (PoC).
    • Affected extension version and Chrome browser version.

Response SLA:

  • Initial Acknowledgment: Within 12 hours.
  • Vulnerability Assessment & Triage: Within 24–48 hours.
  • Patch Release & Web Store Submission: Critical security patches are submitted to the Google Chrome Web Store within 48–72 hours of verification.

🀝 Responsible Disclosure & Recognition

We appreciate the efforts of security researchers who help keep the e-commerce community safe. We are happy to publicly acknowledge researchers who responsibly report verified security findings in our release notes.

There aren't any published security advisories