Enforce trust server-side
Authorization and tenant isolation do not depend on hidden UI controls.
A company’s technology inventory is sensitive operational data. VulnTrust is designed to keep tenant boundaries, authentication, auditability, and recoverability explicit—not implied.
Authorization and tenant isolation do not depend on hidden UI controls.
Security decisions remain tied to their source and application context.
Analysts own workflow and risk decisions; automation supports the process.
We state implemented controls directly and do not claim certifications or independent audits that have not been completed.
Customer-owned API queries resolve through an active organization membership. An organization selector never replaces the server-side check.
Access and rotating refresh JWTs are stored in HttpOnly cookies. Cookie-authenticated mutations require CSRF validation.
Production requires separate Django, JWT-signing, and TOTP-encryption keys supplied outside the codebase.
TOTP enrollment, one-use login challenges, hashed recovery codes, and encrypted TOTP secrets are implemented server-side.
Security-relevant actions can record actor, organization, target, IP address, request ID, user agent, and structured metadata.
Source leases, idempotency keys, durable records, retry controls, and stale-worker recovery reduce duplicate and partial side effects.
The deployed database and Redis are not published as public application endpoints; the gateway exposes only web and API traffic.
AI is not authoritative for vulnerability identity, applicability, severity, finding state, or alert delivery, and provider keys stay server-side.
Uploads are size- and format-limited, originals are not retained, ambiguous rows require review, and committed items preserve source provenance.
Package URL and version queries to public OSV are disabled per organization until an authorized administrator opts in.
VulnTrust separates customer inventory from shared vulnerability intelligence. Customer-specific OSV query coordinates remain only in the tenant-scoped finding evidence that needs them; shared advisory responses do not expose those coordinates. Reports are delivered through authenticated, organization-scoped endpoints.
Contact the VulnTrust team with enough detail for us to understand and respond. Please do not include customer data, credentials, or active exploit material in the first message.
We’ll walk through architecture, tenant isolation, authentication, data flow, deployment controls, and the security work required for your production rollout.