Security & Compliance

Enterprise-Ready Security Controls

This page summarizes the hosted security controls behind Smplify MCP in a way security reviewers and technical buyers can evaluate quickly. It is intentionally lighter than the internal control-mapping documentation and focuses on the controls customers care about during rollout.

Black and white illustration representing security boundaries, verification, and protected access

Hosted transport controls, auditability, tenant-aware design goals, and release hygiene for the Smplify MCP gateway.

Smplify MCP is designed to support common enterprise control requirements, but this page does not claim a formal certification for the MCP gateway by itself. For deeper diligence, use this summary together with the security overview and architecture materials.

What this page covers

Transport and boundary protections

HTTPS enforcement, HSTS, CSP, clickjacking protection, DNS rebinding defense, request-size guards, and browser-safe response headers on the hosted surface.

Authentication and token handling

Per-request bearer-token forwarding, URL-token auth blocked by default, failed-auth throttling, readiness for issuer/audience-bound JWT validation when upstream Smplify auth is ready, and a connector-ready design that keeps any temporary upstream bearer-token bridge inside a dedicated encrypted vault.

Auditability and monitoring

Structured audit events, request correlation IDs, redaction of sensitive values, centralized export, and current Smplify-operated audit sink delivery.

Release and operational assurance

Security regression tests, dependency audit, Bandit, SBOM generation, release provenance workflows, and deployment hardening guidance in the repository.

Control areas supported today

  • Encrypted transport and browser security headers on the hosted gateway
  • Tenant-scoped execution using the caller’s own bearer token
  • Redis-backed rate limiting and repeated failed-auth throttling
  • Audit logging with redaction and request correlation IDs
  • Prompt-injection-aware response sanitization before tool output reaches the model
  • SSRF-safe validation for configured outbound audit webhooks
  • Read-only mode and destructive-operation gating for higher-assurance deployments
  • Connector-ready architecture with explicit-tenant write guardrails and an encrypted vault bridge for future OAuth-connected clients

Enterprise review readiness

Smplify MCP is built to support common enterprise control expectations across areas such as transport security, access control, auditability, change traceability, operational hardening, and release hygiene.

That makes it easier to evaluate during security questionnaires, vendor reviews, and internal AI governance rollouts. When deeper diligence is needed, Smplify can support a formal review with additional architecture and control documentation.

Multi-tenant safety

The hosted gateway is designed so each request runs with the caller’s own bearer token and tenant-scoped permissions. For audit export, Smplify’s long-term design direction is tenant-scoped routing and destination ownership so one customer cannot receive another customer’s events.

That tenant-scoped audit export design is a current architecture priority and is intentionally treated as a productized Smplify capability rather than a customer-managed infrastructure setting.

When deeper detail is needed

Use this page for the public summary, then pair it with the live controls on /security and the safe-usage guidance on /security-best-practices. For deeper diligence, Smplify can also use its internal control mapping and architecture docs during a security review.