Security-first password manager for your business

Passwork is built on a zero knowledge architecture. Security is the foundation of the entire system, not an afterthought. All sensitive data is client-side encrypted before leaving the device. Only your team can access it. No one else.

Trusted by top teams:

Maxon PWC Deutsche Post Orange TDK Victoria Police
Made in EU ISO 27001 certified HackerOne verified security assessment Omdia On the Radar vendor assessment

Why leading companies choose Passwork

  • Made in Europe

    Developed in Europe, with full GDPR and NIS2 compliance and data sovereignty

  • ISO 27001 certified

    Development and infrastructure meet the international benchmark for information security

  • Trusted by public sector

    Chosen by government agencies and highly regulated industries across Europe

  • Enterprise‑grade protection

    Zero-knowledge architecture with client-side encryption keeps your passwords private

  • Independent research shows 30% savings compared to competitors

On-premise solution with flexible deployment

Passwork can be deployed in isolated offline environments and cloud infrastructures to meet corporate security policies and data sovereignty requirements. It runs on Linux and Windows Server, scales horizontally, and integrates with your existing IT stack.

Corporate integration and support

  • LDAP, Active Directory, ALDPro, and SAML SSO authentication
  • SIEM integration for security monitoring
  • MongoDB support included
  • Technical support for deployment and operations
  • Full audit log ownership stays with your organization

Installation flexibility

  • Linux and Windows Server support
  • Docker-based or fully manual component deployment

Scalability and fault tolerance

  • Horizontal scaling
  • High availability via database replication and load balancing

Security built into every stage of development.
Every feature goes through multiple security checks before release.

  • Security by design

    Security requirements defined alongside customer needs before any development begins.

  • Threat analysis

    Threat modeling, security requirements definition, and technical specification planning.

  • Secure coding

    Static code analysis (SAST), dependency checks (SCA), and automated security tests.

  • Security testing

    Dynamic testing (DAST), automated analysis, expert review, and manual validation.

  • Isolated build

    Every build runs in an isolated environment, no external access, no tampering risk.

  • Secure delivery

    Distribution packages are verified and signed to ensure integrity and authenticity.

DevSecOps approach

  • Static and dynamic code analysis
  • Software composition analysis (SCA)
  • AI-assisted review and testing

Secure release process

  • Builds signed via dedicated CI server
  • Offline storage of private keys
  • Annual external penetration testing

Security expertise

  • Security Champions in every team
  • OWASP-based security training
  • Threat modeling for every release

Security standards and independent validation.
Built on recognized industry standards and verified by independent security experts.

ISO-aligned processes

Processes for security and data protection aligned with ISO 27001

Regular security testing

Continuous internal audits and independent external penetration testing

OWASP & SDL compliance

Developed according to OWASP best practices and secure development principles (SDL)

  • Multi-stage code review process
  • No direct commits to the main branch
  • Isolated release build environment

Encryption standards built for enterprise security.
Passwork uses industry-standard cryptography to protect your data at every layer.

  • AES-256

    Encryption for all stored vaults and credentials

  • RSA-2048

    Secure key exchange for data sharing and access control

  • TLS 1.3

    Encrypted channel for all network communication

  • CSPRNG

    Cryptographically secure random number generation

  • OpenSSL

    Industry-standard cryptographic library for data encryption

  • PBKDF2

    Secure key derivation from master passwords

Zero-knowledge architecture.
Your data is protected by a multi-layer encryption system where each level is independently secured.

Choose your plan.
Long-term ownership costs 30% less than the industry average.

  • Standard

    Essential features for small and medium businesses to support secure growth

    3€
    per month /
    per user
    billed annually
    • Quick start with all core features
    • Simple, secure, and low admin overhead
    • Shared vaults, easy access, no training
  • Advanced

    Advanced capabilities for complex security and management needs in large organizations

    4,5€
    per month /
    per user
    billed annually
    • Advanced access management
    • Reliable infrastructure for enterprise environments
    • Priority support and full regulatory compliance

Ready to secure your business?

Join thousands of IT professionals who trust Passwork to manage their passwords securely. Start your free trial today or get a personalized quote.

No credit card required
Full feature access
GDPR compliant
Enterprise-level support

Frequently Asked Questions

Zero Trust means the server holds no information sufficient to decrypt user data. Neither server administrators nor Passwork staff can access passwords — even with full database access.

How this works in practice:

  • • The master password never leaves the user's device
  • • All cryptographic keys are generated on the client
  • • The server stores only encrypted data and encrypted keys
  • • Decryption is only possible on the client side

With client-side encryption enabled, the server cannot read passwords, custom field names and values, TOTP secrets, file contents, or revision history. It can only read non-sensitive metadata: vault names, record titles, login fields, URLs, and tags — fields needed for search and sorting.

Passwork applies two independent encryption layers:

The server layer is always active. All data at rest is encrypted with AES-256-CFB via OpenSSL. This protects against database file theft even without client-side encryption.

The client layer runs when client-side encryption (CSE) is enabled. Sensitive fields are encrypted on the user's device before leaving it. The server receives already-encrypted data and adds a second layer on top.

On Passwork Cloud, CSE is always enabled and cannot be disabled. On on-premise installations, CSE is configurable — disabling it is acceptable for air-gapped environments with documented justification, but is a high-severity finding on any internet-accessible instance.

Each piece of data is protected by a chain where each level encrypts the next:

1. The user enters the master password — it never leaves the device
2. PBKDF2 (SHA-256, 300,000 iterations) derives a 512-bit master key in the browser
3. The master key decrypts the user's private RSA key (2048 bits, OAEP/SHA-256), which is stored on the server in encrypted form
4. The private RSA key decrypts the vault key (256 bits), which is unique per vault and stored RSA-encrypted per user
5. The vault key decrypts the record key (256 bits), unique per record
6. The record key decrypts the password field, custom fields, TOTP secret, and attachment keys
7. Each attached file has its own unique attachment key (256 bits), encrypted with the record key

Compromising a single record key exposes only that one record. It does not expose other records, other vaults, or the user's RSA key.

All algorithms meet or exceed NIST SP 800-131A Rev. 2 recommendations:

  • AES-256 — symmetric encryption (CBC on client, CFB on server via OpenSSL)
  • RSA-2048 OAEP / SHA-256 — asymmetric key exchange between users; generated via WebCrypto API
  • PBKDF2 — key derivation from master password: 300,000 iterations (SHA-256) on the client, 600,000 iterations (SHA-512) on the server for login password hashing
  • TLS 1.3 — all network connections; TLS 1.0 and 1.1 are disabled
  • CSPRNG — all keys, IVs, and tokens use cryptographically secure generators: WebCrypto in the browser, OS-level generators on the server (getrandom() on Linux, CryptGenRandom on Windows)
  • PBKDF2 salts — 20 characters, ~120 bits entropy, unique per user, server-generated

Each key uses a fresh 128-bit initialization vector per encryption operation. Vault and record keys carry ~596 bits of input entropy, reducing to 256 bits through the AES key derivation step.

  • ISO 27001 certified (2024) — covers development, infrastructure, and operational practices
  • GDPR compliant — data minimization, retention controls, right to deletion, data portability
  • Tested by HackerOne — independent penetration testing covering OWASP Top 10, SANS Top 25, authentication flows, API authorization, management, and resilience against advanced persistent threats
  • NIST SP 800-131A Rev. 2 — all cryptographic algorithms comply; AES-256 and RSA-2048 are approved; PBKDF2 iteration counts are within NIST guidance

Passwork also publishes a machine-readable security profile at passwork.pro/trust.json and a standard security.txt contact file.

Every feature goes through six mandatory stages before release:

1. Idea stage — security requirements analysis and Security Champion review
2. Analysis — STRIDE threat modeling and specification
3. Code — SAST (static analysis), SBOM checks on all dependencies, automated tests
4. Build — isolated CI environment, signed distributives, private keys stored offline
5. Testing — DAST (dynamic analysis), AI-assisted review, manual expert verification
6. Release — signature verification on the customer portal

All distributives are cryptographically signed. Builds are assembled in an isolated CI environment with no external access. Private signing keys are stored offline. The public key is available at passwork.pro/public-key. Signature verification is built into the deployment scripts, so installations automatically confirm integrity before proceeding.

Passwork supports multiple authentication layers:

  • Two-factor authentication (TOTP) — Google Authenticator, Microsoft Authenticator, or the Passwork 2FA app. For administrative accounts, physical FIDO2/WebAuthn security keys are the recommended option — they are phishing-resistant, unlike TOTP codes
  • SSO — SAML 2.0, Azure AD (Microsoft Entra ID), Keycloak, AD FS
  • LDAP / Active Directory — with LDAPS (port 636, TLS) required; the LDAP service account must be read-only
  • Account lockout — configurable; the recommended setting is lockout after 5 failed attempts with a 15-minute cooldown, in line with NIST SP 800-63B
  • Session management — access tokens (~2.8-hour lifetime), refresh tokens (36-hour lifetime), configurable inactivity timeout, CSRF protection, and cookie Secure + SameSite flags
  • Role-based access control — 2FA enforcement is configurable per role; administrators can terminate any user's active sessions from the admin panel

All user and system actions are recorded in an immutable action history. The log captures authentication events, admin actions, vault access changes, password exports, API token creation, and LDAP sync results. Events are exported in CEF (Common Event Format) via Syslog, compatible with Splunk, IBM QRadar, Microsoft Sentinel, Elastic SIEM, and ArcSight. On Windows Server, events are written to Windows Event Viewer.

Recommended retention:

  • • General use: 90 days minimum
  • • ISO 27001: 1 year
  • • PCI-DSS: 1 year, with 3 months immediately available

The Security Dashboard continuously analyzes vault credentials and flags weak passwords, passwords not rotated in over 180 days, and credentials that were accessed by users whose access has since been revoked — the last category is a critical finding requiring immediate rotation.

Send a report to [email protected]. Passwork will acknowledge receipt within 5 business days and coordinate on disclosure timing (typically 30–90 days after a fix is released). Scope includes Passwork Cloud, on-premise installations, browser extensions, mobile apps, API, and documentation. Researchers who follow the responsible disclosure policy have safe harbor — no legal action for good-faith testing. Researchers credited with confirmed findings are listed in the public Hall of Fame.

Got any questions? — Help center