Skip to main content

Application security

Centrifuge applications are built to meet institutional requirements for access control, infrastructure isolation, and real-time monitoring.

Access controls​

  • All administrative access requires hardware-based two-factor authentication. Hardware security keys are phishing-resistant and prevent the most common account hijacking attacks.
  • Sensitive operations such as deployments and configuration changes require multi-signature authorization, preventing any single point of failure.
  • Application deployments are authenticated with strict role-based access controls. Infrastructure platform access is limited under least-privilege principles.

Infrastructure​

  • Centrifuge web applications run on a fully serverless architecture, minimizing the attack surface by eliminating persistent servers.
  • Containerized services undergo security scans before publication, with regular base image and dependency updates.
  • Static code analysis runs across the development pipeline, catching vulnerabilities before they reach production.

Network and application protection​

All web traffic is proxied through an enterprise CDN and security platform, providing layered protection from network edge to application layer:

  • Full strict SSL/TLS with TLS 1.3, automatic HTTPS redirection, and HSTS enforcement to prevent downgrade attacks.
  • DNSSEC enabled on all managed zones, preventing DNS spoofing and cache poisoning.
  • Web application firewall with managed rulesets, DDoS mitigation, and rate limiting.
  • Strict Content Security Policy, Permissions-Policy, and other security headers enforced across all applications.
  • Client-side script monitoring to detect tampering.
  • Fail-closed architecture: if edge services hit their limits, requests are rejected rather than bypassing security controls.
  • Hardware-based two-factor authentication enforced for all platform account users.

Continuous security monitoring​

Security monitoring spans multiple layers: cloud infrastructure, CDN, application code, and onchain activity. Using a combination of managed security platforms and automated scanners, in many cases multiple layers and scanners are in place at different stages.

Includes:

  • Static application security testing (SAST) and source code scans.
  • Vulnerability detection and license compliance checks across all open-source dependencies.
  • Known vulnerability scanning for all published container images.
  • Cloud infrastructure configuration audits and posture management.
  • Surface monitoring for public-facing domains.
  • Infrastructure-as-code misconfiguration detection before deployment.

Onchain monitoring​

Onchain activity is monitored in real time through Hypernative, alongside our own dedicated monitoring tools.

  • Hypernative watches the deployed protocol contracts across all supported networks, flagging anomalous transactions, permission changes, and state changes as they happen.
  • Detections route to the on-call security rotation, so the guardian multisigs can pause the protocol if an incident requires it.
  • Dedicated monitoring tools track sensitive onchain operations: liquidity pool activity, cross-chain message relaying, and code deployment verification against published releases.