Skip to main content

SECURITY ARCHITECTURE

The Architecture Behind the Guarantee

We do not ask you to trust us. We ask you to verify the mathematics. Every cryptographic primitive is NIST-standardised, open-source, and auditable directly from your browser console — no proprietary black boxes.

PROOF & DOCUMENTATION

Everything in One Place

A reviewer should never have to hunt for proof. Architecture, scope, threat model, disclosure policy, and our open-source roadmap — linked directly below.

Status

Audit Status

Internal review complete, bug bounty active, external audit planned (not yet scheduled).

Document

Security Whitepaper

Full cryptographic protocol — AES-256-GCM, random keys, Shamir SSS, heir release.

Policy

Privacy Model

Zero PII stored. Data minimisation by design. GDPR-aligned.

On this page

Threat Model

Six adversary classes, impact, and mitigations.

On this page

Audit Scope

Complete surface published before engagement — no exclusions.

Policy

Responsible Disclosure

Rules of engagement, safe-harbour, and how to report.

Programme

Bug Bounty Programme

Severity tiers, scope, Hall of Fame + CVE credit.

RFC 9116

security.txt

RFC 9116 machine-readable disclosure contact.

GitHub

Open-Source Crypto-Core

AES-256-GCM · SHA-256 · Shamir SSS · MIT — live on GitHub.

Internal Security ReviewComplete · 2026-06-14

Grey-box source review of the crypto core, all API routes, auth/MFA, and the succession protocol, modelled on Cure53 methodology. No Critical or High findings. All Medium findings resolved before public launch; residual items are Low/Info hardening, tracked internally. External professional audit planned — scope ready, auditor not yet selected and no date committed; full report will be published on completion.

Download internal review (PDF) →

ZERO-KNOWLEDGE DATA FLOW

How Your Data Actually Moves

Your Owner Key (Shard 1) is downloaded to your device and never transmitted to VaultPass. We hold at most one of the remaining shards — never enough to reconstruct your vault alone. Every encryption step runs in your browser before any data is sent.

01 · INPUT••••••••Seed Phrasetyped in browserYour device onlyPLAINTEXT BOUNDARY02 · ENCRYPTIONAES-256-GCMWeb Crypto APINEVER LEAVES DEVICERandom 256-bit key · Shamir 2-of-303 · SHARDINGShamir's 2-of-3 SSSSHARD 1Owner · You hold thisSHARD 2Heir · Time-lockedSHARD 3Guardian Node · Cold storageAny 2 shards reconstruct your vault · 1 alone reveals nothing

ALL THREE STEPS OCCUR IN YOUR BROWSER — NO PLAINTEXT IS EVER TRANSMITTED

3-STEP ZERO-KNOWLEDGE FLOW

How VaultPass Stays Zero-Knowledge

From the moment you type your seed phrase to the moment shards reach their guardians — every transformation happens on your device. Watch it happen below.

Prove it yourself →
SEED PHRASE
•••• •••• ••••
● SENTINEL ACTIVE
SHARD 1
Owner Vault
SHARD 2
Heir Vault
SHARD 3
Guardian Vault
SHARD 1
Owner Vault
SHARD 2
Heir Vault
SHARD 3
Guardian Vault
ZERO-KNOWLEDGE VERIFIED
No server holds your plaintext
STEP 01· INPUT & SHIELD
Client-Side Encryption Initialized
Your seed phrase never leaves your device unencrypted

DISTRIBUTED VAULTING · INTERACTIVE

Sentinel Sharding — Live

Distribute your vault key across three independent guardians, then reconstruct it using any two. No single guardian holds enough to open the vault — not even us.

SHARD 1
Owner · You
·
SHARD 2
Heir · Time-locked
·
SHARD 3
Guardian Node · Cold
·
VAULT SECURED
STEP 1 · Distribute your vault key across 3 independent guardians

SHAMIR'S SECRET SHARING · 2-OF-3 THRESHOLD · OPEN STANDARD

THE SENTINEL PROTOCOL

How the Dead-Man's Switch Works

“Security is not a promise; it is a verifiable sequence of cryptographic events.”

01ENCRYPTION

Zero-Knowledge Encryption

Before any data is stored, it is encrypted on your device using AES-256-GCM — a NIST-standardized authenticated-encryption cipher. Your master key never touches our servers.

Trust FactorAES-256-GCM is NIST-standardised authenticated encryption. The authentication tag makes any tampering mathematically detectable.
02MONITORING

The Heartbeat (Sentinel Pulse)

You define your inactivity threshold: 30, 180, or 365 days. A secure Cloud Function monitors the timestamp of your last check-in. This is the only metadata we track.

Trust FactorThe only data logged is a UTC timestamp of your last check-in. No IP address, no device fingerprint, no behavioural data is stored.
03VERIFICATION

The Verification Phase (Grace Period)

A missed heartbeat does not immediately trigger release. The system enters a Grace Period and sends encrypted notifications via Email. You can reset the timer at any point with a single click.

Trust FactorThree warning emails are sent at 80%, 90%, and 100% of the interval. Each contains a one-click cancel link. No technical skill required to stop the trigger.
04RELEASE

The Event Horizon (Inheritance Release)

Once threshold + grace period expires, the Sentinel releases its shard. Combined with the heir's sealed shard, the vault reconstructs — without your Owner Key, without VaultPass staff involvement, and without any central point of control.

Trust FactorThe heir's token is single-use and SHA-256 hashed on our end. Shard recombination uses the public Shamir SSS algorithm and runs entirely in the heir's browser. Your Owner Key (Shard 1) never leaves your device.

OPEN SOURCE PROMISE

The cryptographic primitives are browser-native and independently verifiable today. The crypto-core module — AES-256-GCM, SHA-256, and Shamir's Secret Sharing — is open-source and MIT-licensed on GitHub right now. The crypto path is left unobfuscated in the shipped app, so you can read it in your browser's DevTools and confirm it matches the public source — no hidden cryptography. That verifies the algorithm itself; the independent external audit, which covers the whole application, is planned — scope finalised, not yet scheduled.

AES-256-GCMVault encryption

NIST-standardised authenticated encryption. Browser-native via Web Crypto API.

crypto.subtle.generateKey({name:"AES-GCM",length:256},true,["encrypt","decrypt"])
CSPRNGVault key generation

Each vault key is a fresh random 256-bit key from the browser CSPRNG — not derived from a password.

crypto.getRandomValues(new Uint8Array(32))
Shamir's Secret SharingKey sharding

Splits your key into N shards; any M can reconstruct it. Mathematically proven; no central custodian.

github.com/lionzion351-dev/crypto-core (open source · live)

Open any browser DevTools console on this site and run the snippets above. The cryptography is real.

DESIGN RATIONALE · PRIMITIVE SELECTION

Why AES-256-GCM and not ChaCha20-Poly1305?

Both are AEAD (Authenticated Encryption with Associated Data) ciphers with equivalent security margins. AES-256-GCM is browser-native via the Web Crypto API, which eliminates the risk of a flawed JavaScript implementation of ChaCha20. Hardware AES acceleration is available in every modern CPU — latency is negligible. The authentication tag (GHASH) ensures any ciphertext modification is detected before decryption.

Why Shamir's Secret Sharing over threshold signatures (e.g. FROST)?

Shamir SSS operates on a finite field (GF(2^8) or GF(prime)) using polynomial interpolation — it is information-theoretically secure. A shard below the threshold reveals absolutely zero bits of the secret, provably. Threshold signature schemes like FROST require all participants to be online simultaneously for key generation. SSS allows offline shard distribution — critical for an inheritance protocol where heirs may only be online months or years after setup.

Why isn't the vault key derived from my password?

Because a password-derived key would require your heir to know your password — which defeats the entire point of password-free inheritance. Instead, each vault uses a fresh random 256-bit key from the browser CSPRNG, split via Shamir's Secret Sharing so any 2 of 3 shards reconstruct it. Your heir recovers the vault from shards, never from a password you'd have to share. (PBKDF2 key-stretching is used only in the standalone importer that decrypts files you bring from other password managers.)

Open Source · Live

github.com/lionzion351-dev/crypto-core
AES-256-GCM · SHA-256 · Shamir SSS implementation · MIT licensed

Verify on GitHub →

LIMITATIONS & RISK DISCLOSURE

No security product is absolute. Here is exactly what VaultPass does not protect against, and how your secrets could still be lost — so you can plan around it.

It is not a full estate plan or legal advice

VaultPass delivers your encrypted secrets to a chosen heir. It does not replace a will, an executor, or legal counsel. Treat it as one layer alongside your hardware-wallet backups and legal documents.

Your heir must act — correctly and in time

Release is automated; recovery is not. If your heir ignores notifications, loses their access, or cannot follow the recovery steps, the vault is not opened for them. Brief your heir in advance and keep their contact details current.

Secrets can be lost if enough shards are lost

The vault key is split 2-of-3. If you lose your owner shard and the server-held shards also become unavailable, the vault cannot be reconstructed. There is no master backdoor — that is the point, and the risk. Keep your owner shard and an encrypted export safe.

Email is part of the recovery chain

Check-in reminders and heir notifications use email. A compromised or abandoned inbox (phishing, SIM-swap on any SMS fallback, provider lockout) weakens that link. Use a secure, monitored inbox and enable MFA.

Trust boundary while VaultPass operates

While the service runs, VaultPass holds the encrypted heir and sentinel shards and combines them only when your switch fires. A full internal compromise of both server-held shards after a trigger is the worst case we design against — and a key reason the independent external audit matters. Your plaintext and owner shard are never on our servers.

Recovery if VaultPass disappears

The reconstruction math is open-source (crypto-core). Exported shards can be recombined without our servers, so a permanent shutdown does not strand a vault you have exported. Keep an encrypted export.

AUDIT ROADMAP

Phase 01Internal Security Review
CompleteQ1 2026

Full threat-model analysis, code review, and cryptographic protocol verification by the core team.

Phase 02Community Bug Bounty
ActiveActive now

Public invitation for independent researchers to find and responsibly disclose vulnerabilities. Valid findings earn Hall of Fame credit and CVE recognition (no cash bounties).

Phase 03External Professional Audit
PlannedPlanned · not yet scheduled

Independent third-party cryptographic audit by a specialist security firm. Full report will be published publicly upon completion.

COMPLIANCE & CERTIFICATIONS

Institutional Standards

Commitment to regulatory and industry frameworks is not a checkbox — it is a constraint that shapes architecture decisions from day one.

SOC 2 Type IITrust Services CriteriaIn Progress

External audit planned — scope ready, not yet scheduled. Full report published on completion.

GDPREU / EEA Data ProtectionCompliant

Zero PII stored. Data minimisation by design. No IP logging.

NIST CSF 2.0Cybersecurity FrameworkAligned

Identify · Protect · Detect · Respond · Recover

EXTERNAL AUDIT SCOPE

Published Scope Document

Published prior to engagement so any researcher — not just our auditor — can review the same surface area. Scope is intentionally complete: no exclusions that would hide known weaknesses.

Auditor not yet selected · Scope finalised and ready · Engagement begins once scheduled

Cryptographic Implementation

AES-256-GCM encryption / decryption correctness
Shard encryption at rest (AES-256-GCM, HMAC-derived keys)
Shamir's Secret Sharing polynomial math
Random number generation via Web Crypto API
IV / nonce uniqueness and reuse prevention

Authentication & Session Management

Supabase JWT issuance and validation
Session token lifetime and revocation
Account lockout and brute-force resistance
OAuth flow security (PKCE, state param)
MFA implementation correctness

API & Server Security

Input validation and output encoding on all endpoints
Rate limiting effectiveness (Upstash Redis)
Authorisation checks and RLS bypass attempts
SSRF, SSTI, injection vectors
Stripe webhook signature verification

Client-Side Security

Content Security Policy headers
XSS vectors in vault rendering
Clipboard and memory clearing after session
Browser extension isolation model
Local storage / sessionStorage usage

Heir Delivery Protocol

Shard release trigger logic correctness
Heir token single-use enforcement
Grace period notification timing
Shard recombination in heir browser
Replay attack prevention

Infrastructure & Supply Chain

Dependency vulnerability audit (npm audit)
Vercel deployment configuration review
Supabase RLS policy completeness
Secret scanning and rotation practices
Third-party library integrity

Full audit report, including findings, severities, and retest confirmation, will be published publicly at this URL upon completion. No findings will be withheld. Auditing firm to be announced on engagement.

TRUST TIMELINE

Verifiable History

Every security milestone, audit event, and infrastructure change — timestamped and permanent. Updated as events occur; nothing retroactively edited.

Q4 2025
Zero-Knowledge Protocol Specified

Architecture design complete. Zero-knowledge constraint formalised: plaintext must never leave the client device. Threat model drafted.

Q1 2026
Internal Security Review CompleteAudit

Full threat-model analysis, cryptographic protocol review, and code audit by the core team. All critical findings resolved before public launch.

Q2 2026
Public Beta Launch

VaultPass opened to early users. Random-key AES-256-GCM encryption and Shamir 2-of-3 sharding live in production.

Q2 2026
Responsible Disclosure Programme Launched

Public vulnerability disclosure programme opened. Valid findings earn Hall of Fame credit and CVE recognition. Responsible disclosure policy published at /security.

Q2 2026
Crypto-Core Open-Sourced

AES-256-GCM, SHA-256, and Shamir's Secret Sharing implementation published as a standalone, dependency-free, MIT-licensed module on GitHub — ahead of the external audit, for independent review.

Planned
External Professional Audit — PlannedPlanned

Independent third-party cryptographic and application security audit planned. Scope finalised and ready; auditor not yet selected and no date committed. Full public report to be published upon completion.

2027
SOC 2 Type II — In ProgressPlanned

External SOC 2 Type II audit covering Security, Availability, and Confidentiality Trust Services Criteria. Report to be published publicly.

RESEARCHER ACKNOWLEDGEMENTS

Hall of Fame

Security researchers who disclosed vulnerabilities responsibly and helped make VaultPass safer. Listed with permission. We are bootstrapped and offer recognition and CVE credit, not cash payments.

NO VULNERABILITIES REPORTED YET

Be the first. Valid findings earn public recognition here and CVE credit. We are bootstrapped — we offer acknowledgement, not cash. All reports acknowledged within 48 hours.

[email protected]

REFERENCED IN /.WELL-KNOWN/SECURITY.TXT · RFC 9116 COMPLIANT

ARCHITECTURE TEAM

Two Domain Specialists. No Generalists.

Cryptographic decisions and product decisions are handled by separate, focused roles. No single person controls both the key architecture and the user-facing protocol — by design.

ROLE
Security Architect
Cryptographic Protocol Design
AES-256-GCM implementation
Random key generation (CSPRNG)
Shamir Secret Sharing
Threat modelling & zero-trust design

Every cryptographic primitive is NIST-standardised. Every integration is peer-reviewed before deployment. The architecture is designed so any competent cryptographer can verify it without taking our word.

ROLE
Product Lead
Protocol Architecture & Succession UX
Dead man's switch design
Heir notification protocol
Zero-friction legacy access
Regulatory compliance framework

The inheritance trigger must work correctly the first time — for a person who has just lost someone. That single constraint shapes every product decision: no friction, no dependency on VaultPass staff, no lawyers required.

ROLE
External Security Review
Independent Protocol Oversight
Responsible disclosure programme — active
Researcher Hall of Fame — public
Third-party cryptographic audit — planned
Full public report on completion

We are confident in our architecture — confident enough to invite independent researchers to break it. No security claim at VaultPass is considered final until it has survived external scrutiny.

THREAT MODEL

What We're Defending Against

A security architecture is only meaningful when measured against specific, named threats. Below is the complete threat model for VaultPass — including threats we cannot fully prevent and why zero-knowledge limits their blast radius.

Compromised VaultPass InfrastructureNation-state / insider / supply chain
Contained

Impact if successful

Database read access — ciphertext only

Mitigation

Zero-knowledge by design. AES-256-GCM ciphertext without the master key is computationally indistinguishable from random bits. 2^256 keyspace — brute force is physically impossible within the age of the universe.

Network Interception (MITM)ISP / rogue certificate / adversarial hotspot
Contained

Impact if successful

Traffic inspection or credential intercept

Mitigation

TLS 1.3 enforced on all endpoints. All plaintext encryption occurs client-side before any byte leaves the device. A successful MITM yields only ciphertext and salted auth tokens.

Compromised User DeviceMalware / keylogger / rogue browser extension
Residual Risk

Impact if successful

Seed phrase capture at point of entry

Mitigation

Session inactivity lock. No persistent plaintext in memory between sessions. Vault contents cleared on tab close. No clipboard persistence. 2FA enforced on account login.

Weak or Reused Account PasswordCredential stuffing / dictionary attack
User-Controlled

Impact if successful

Account takeover via login

Mitigation

Supabase bcrypt-hashed auth credentials. Upstash Redis rate limiting (5 req/min on auth endpoints). Account lockout after 10 failed attempts. The vault key is a random key, not derived from your password, so a cracked login password alone does not decrypt vault ciphertext without the owner's separately-held shard.

Social Engineering of VaultPass StaffPretexting / coercion / insider threat
Contained

Impact if successful

Unauthorized access to production systems

Mitigation

No employee accesses vault plaintext in normal operation, and no interactive decryption back door exists. Production access requires MFA and is fully audited, with separation of duties between infrastructure and cryptographic roles.

Cryptographic Weakness / Quantum AttackCryptanalyst / large-scale quantum computer
Theoretical

Impact if successful

Theoretical weakening of encryption guarantees

Mitigation

AES-256 is quantum-resistant to Grover's algorithm (effective 128-bit post-quantum security). SHA-256 resistant to known quantum attacks. Post-quantum migration path (ML-KEM / CRYSTALS-Kyber) planned within the external audit scope.

INCIDENT RESPONSE POLICY

How We Respond When Things Go Wrong

Detection → Triage → Containment → Remediation → Disclosure. Every incident is assigned a severity tier with binding response time commitments. Affected users are notified before the public.

P0Critical

Examples

Confirmed data breach or unauthorised data access
Cryptographic implementation flaw
Authentication bypass enabling account takeover

Acknowledge

< 1 hour

Notify Users

< 4 hours

Remediate

< 24 hours

Disclose

Within 72 hours

P1High

Examples

Service degradation > 10 minutes
Sentinel heartbeat trigger failure
Payment processing outage

Acknowledge

< 4 hours

Notify Users

< 8 hours

Remediate

< 72 hours

Disclose

Within 7 days

P2Medium

Examples

Intermittent performance degradation
Non-critical feature unavailable
Third-party service disruption

Acknowledge

< 24 hours

Notify Users

< 48 hours

Remediate

Next sprint

Disclose

On resolution

Report incidents: [email protected] · Status updates posted to vaultpass.network/status · Post-mortems published for all P0 and P1 incidents.

VULNERABILITY DISCLOSURE POLICY

Responsible Disclosure — Full Policy

We operate a public responsible disclosure programme. Researchers who follow this policy will not face legal action. Valid findings earn public credit and Hall of Fame acknowledgement. We respond within 48 hours and target a 90-day patch window.

CriticalCVE + Hall of Fame
Authentication bypass
Remote code execution
Cryptographic flaw enabling vault decryption
Mass data exposure
HighHall of Fame
Sensitive data exposure affecting other users
SSRF
Significant privilege escalation
MediumHall of Fame
CSRF on authenticated actions
Stored XSS
Insecure direct object reference
LowPublic credit
Security header misconfiguration
Verbose error messages leaking stack traces
Best-practice violations without direct impact

In Scope

vaultpass.network and all subdomains
VaultPass API endpoints
iOS mobile application
Browser extension
Authentication and session management

Out of Scope

Volumetric denial-of-service attacks
Social engineering of VaultPass staff
Physical attacks against infrastructure
Vulnerabilities in third-party services
Reports already known to us

Safe Harbor · Security research conducted in good faith under this policy will not result in legal action. We will not refer researchers to law enforcement for responsible disclosure that adheres to these guidelines. Your research helps make VaultPass safer for everyone.

SECURE SOFTWARE DEVELOPMENT LIFECYCLE

Security at Every Stage of Development

Security is not a review step at the end of a sprint. Every phase of the VaultPass development lifecycle has security controls built in — from whiteboard to production monitoring.

Design

Threat model required before implementing any security-sensitive feature
Cryptographic primitive selection reviewed by security architect
Zero-knowledge constraint enforced at architecture level — plaintext must never leave the client
Privacy impact assessment for any feature touching user data

Development

TypeScript strict mode — no implicit any, exhaustive null checks enforced
ESLint security rules: no dynamic code execution, no raw HTML injection, no untrusted DOM writes
Dependencies pinned and scanned via npm audit on every CI run
Cryptographic code: Web Crypto API only — no third-party cryptographic JS libraries

Review

All changes require peer review and approval before merge to main
Security-sensitive paths (crypto, auth, payments) require security architect sign-off
No direct commits to the main branch — all changes via pull request
Automated SAST (static analysis) scan triggered on every pull request

Deploy

CI/CD via Vercel — no manual SSH production deployments
Secrets injected at build time via environment variables; never committed to source
Preview deployments for every pull request — reviewed before production promotion
Automatic rollback triggered on error rate spike above threshold

Monitor

Upstash Redis rate limiting on all auth, payment, and vault API endpoints
Health check endpoint probed every 60 seconds — alert on failure
Anomaly and error alerts via Inngest + Telegram operations channel
GitHub Dependabot alerts for dependency vulnerabilities reviewed within 24 hours

KEY MANAGEMENT PRACTICES

Complete Key Lifecycle

The security of your vault is entirely determined by how its encryption key is generated, stored, used, and destroyed. Every stage of that lifecycle is documented here.

Generation

All cryptographic keys generated using the browser CSPRNG via Web Crypto API (window.crypto.getRandomValues). No server-side key generation for user vault keys. Randomness cannot be influenced or observed by VaultPass.

Key Model

VaultPass does not derive vault keys from a password. The random per-vault key is used directly with AES-256-GCM and then split via Shamir's Secret Sharing — there is no master-password KDF in the vault path.

Storage

The full vault key is never persisted. Only AES-256-GCM ciphertext and encrypted key shards are stored in Supabase — a database-only compromise yields ciphertext, not plaintext.

Sharding

Vault key split via Shamir's Secret Sharing (2-of-3). Owner shard downloaded to your device and never uploaded; heir and sentinel shards stored by VaultPass, encrypted at rest with secrets held outside the database. Any single shard reveals zero bits. The heir + sentinel shards are used to reconstruct the key during heir release.

Rotation

Re-encrypting your vault generates a fresh random vault key, re-splits it via Shamir, and replaces all stored ciphertext and shards browser-side. Previous ciphertext is overwritten and irrecoverable.

Destruction

Account deletion triggers immediate removal of all ciphertext, Shamir shards, and associated metadata from Supabase. No soft-delete or recovery window for vault contents. Destruction is permanent and irreversible.

EMPLOYEE ACCESS CONTROLS

Internal Access Policy

Your vault is encrypted on your device before it reaches us — we store ciphertext, not plaintext. The encrypted heir and sentinel shards we hold for automated delivery are combined only when the dead man's switch fires. The controls below govern everything else.

Principle of Least Privilege

All internal service accounts scoped to minimum required permissions. Supabase Row-Level Security (RLS) enforced on every table — no application-layer bypass possible. Service role keys used only for server-side operations that explicitly require them.

Zero-Knowledge Enforcement

No VaultPass employee accesses user vault plaintext in normal operation, and the full vault key is never transmitted to our servers — owner shards stay on user devices. Note that automated heir release requires our infrastructure to hold the heir and sentinel shards, so this is end-to-end encryption with automated delivery rather than a technical inability to decrypt.

Multi-Factor Authentication

MFA required for all production system access: Supabase dashboard, Vercel deployment, GitHub, and Stripe. TOTP authenticator apps required — SMS MFA explicitly prohibited due to SIM-swap risk.

Audit Logging

All production database access, deployment events, and admin actions are logged with timestamp, identity, and action. Logs are immutable and retained for 12 months. Anomalous access patterns trigger automated alerts.

Code & Repository Access

Repository access restricted to active team members. All commits GPG-signed. Main branch protected — no direct pushes. Automated secret scanning on every push prevents accidental credential commits.

Offboarding Protocol

Departing team members have all access revoked within 24 hours across GitHub, Supabase, Vercel, Stripe, and shared credentials. OAuth tokens rotated immediately. Quarterly access review performed for all active accounts.

RECOVERY GUARANTEES

What We Guarantee, In Writing

Most password managers ask you to trust their uptime. VaultPass is designed so that even a full service failure cannot lock you or your heirs out of your vault. These are not aspirational targets — they are architectural constraints.

Inheritance Without VaultPass

Because your Owner shard lives on your device and the Heir shard is held by your designated heir, vault reconstruction requires no VaultPass infrastructure. Any 2-of-3 Shamir shards reconstruct the vault key. Your inheritance plan works even if VaultPass ceases to exist.

Data Durability

Encrypted vault data stored in Supabase with continuous replication across multiple EU availability zones. Point-in-time recovery available for 30 days. Target RPO: 0 seconds (synchronous replication). Target RTO: < 15 minutes for full restoration.

No Vendor Lock-in

Your encrypted vault export is available at any time from the dashboard in standard JSON. The export is AES-256-GCM encrypted and readable with the open-source crypto-core module (live on GitHub) — no VaultPass software required to decrypt your own data.

Service Continuity Commitment

If VaultPass is acquired, merged, or ceases operations: 90-day notice to all users; full vault export delivery to all active accounts; open-sourcing the full cryptographic core before shutdown so heirs can reconstruct vaults independently forever.

Uptime SLA Target

99.9% monthly uptime target for API and Heartbeat Sentinel (< 44 minutes downtime/month). Planned maintenance announced 48 hours in advance. Sentinel trigger clock pauses during verified VaultPass outages — your inactivity window is never unfairly consumed.

Export your encrypted vault at any time from the Dashboard → Settings → Export. Your data is always yours.

END-TO-END SECURITY

Your Email is Part of the Security Chain

VaultPass protects your vault with AES-256-GCM encryption and zero-knowledge architecture. But the last mile — delivering your vault to your heir — depends on email. A compromised or inaccessible inbox can break the chain. Here is what we recommend.

Your Email (Vault Owner)

🔐

Strong unique password

Use a password manager for your email account. Never reuse passwords across services.

📱

Enable 2FA

Use Google Authenticator or Microsoft Authenticator — not SMS, which is vulnerable to SIM-swap attacks.

🔁

Recovery method

Add a backup email and phone number. Losing email access means losing check-in ability, which could trigger your vault prematurely.

⚠️

Your responsibility

VaultPass sends check-in reminders and warning emails to this address. Losing access to it could trigger your dead man's switch unintentionally.

Your Heir's Email

📬

Active & accessible inbox

The vault transfer token is delivered by email. If your heir cannot access their inbox, they cannot claim the vault.

📱

Enable 2FA on their account

A compromised heir inbox means an attacker could intercept the vault transfer. Authenticator-based 2FA prevents this.

🛡️

Safe-sender list

Ask your heir to add vaultpass.network to their trusted senders now — before it matters — so transfer emails are never caught by spam filters.

🔁

Brief them now

Don't wait until the switch triggers. Tell your heir about VaultPass, their role, and that they should keep this email address active.

Our commitment: VaultPass secures what we control — the vault encryption, the key sharding, the trigger protocol. What we cannot control is whether you and your heir maintain access to your email accounts. That responsibility is yours. We strongly recommend briefing your heir when you set up your vault, not after you need it.

RESPONSIBLE DISCLOSURE

Found a Vulnerability?

We are confident in our code — confident enough to invite researchers to break it. Valid findings earn public acknowledgement in our Hall of Fame. We are a bootstrapped startup: we offer recognition, credit, and a CVE, not cash. Disclosure acknowledged within 48 hours.