The platform that assesses your environment must also protect it.
Authorization, tenant isolation, evidence integrity and approval boundaries are first-class system constraints — not add-ons added after the scanner works.
Verified authorization
Active security work is gated behind ownership verification, current authorization and explicit scan scope.
Tenant isolation
Organization context is enforced across application and database layers, with row-level security as a core boundary.
Immutable approvals
Remediation approval is bound to an exact plan version and hash. Material changes require a fresh decision.
Restricted private agent
The private agent is outbound-only and capability constrained. It does not expose an arbitrary remote shell.
Traceable evidence
Findings retain scanner provenance and sensitive actions are designed to leave an auditable trail.
Constrained AI
AI analyzes normalized, redacted evidence and proposes actions. It does not authorize scans or arbitrary commands.
Recommendation and authority are deliberately separate.
AI can interpret evidence and propose a typed remediation plan. Approval is recorded against the exact plan version and hash. Execution performs preflight checks, controlled actions and verification; arbitrary shell or script payloads are rejected by the plan contract.
Explore governed remediationFound a security issue in TENSISSEC?
Please send a concise report to security@tensissec.com, including the affected surface, reproduction information and impact. Do not test against customer environments or access data beyond what is necessary to demonstrate the issue.
For general product support, use support@tensissec.com.
Security assurance is continuous
TENSISSEC combines implementation controls, automated regression gates, pinned security runtimes, controlled validation targets, audit trails and release discipline so security is evaluated throughout the product lifecycle rather than treated as a one-time launch exercise.
View validation research →