Implement Privacy by Design Principles

Privacy by Design guides the incorporation of personal-data protections into software architecture and implementation. It applies to databases, APIs, authentication, and user flows, using Privacy by Design principles and GDPR Article 25 as references. CCPA and LGPD are also named frameworks to consider for the users and regions a product targets.

The core principles are data minimization, purpose limitation, storage limitation, privacy-preserving defaults, end-to-end security, and transparency. Each collected field needs a documented reason. Purpose and retention belong in the schema or metadata, and expired data should be deleted or anonymized through an automated policy rather than retained indefinitely.

Optional collection is opt-in, with sensitive analytics or marketing settings off by default. Security guidance includes encryption at rest and in transit, role-based access, and audit records of sensitive-data access. Transparency includes clear explanations of collection and accessible user controls.

User-facing rights flows cover returning all user data, correcting records, account deletion and data purging, and machine-readable export such as JSON or CSV. Backup retention belongs in the design as well; deleting an active record without considering retained backups leaves part of the data lifecycle unaddressed.

Use this skill when building applications that collect names, email addresses, locations, preferences, or other personal data. It also applies to forms, accounts, analytics, third-party services, and pre-launch checks of privacy controls.

The guidance discourages collecting fields just in case, loading tracking before consent, exposing internal errors, and postponing deletion or export design. Its code patterns are implementation aids and require validation in the actual environment rather than serving as a substitute for testing or expert review.

Inputs include proposed data fields, documented collection purposes, consent choices, storage lifetimes, API response requirements, dependency behavior, and user-control needs. Outputs include minimized models, purpose and retention annotations, consent checks, response allowlists, safe logs, and feature-level audit checklists.

Consent-gated tracking can use this pattern:

if (userConsent.analytics) {
  analytics.track(userId, event);
}

Other examples contrast a broad user object with a minimal model, plain-text personal-data logging with hashed identifiers, and full-object API responses with explicit allowlists. Generic client errors avoid exposing internal details, while structured server logging needs redaction and allowed fields.

Before adding a data-touching dependency, the audit asks what it collects, where it sends data, when it loads relative to consent, whether collection can be disabled, and whether its policy aligns with the application. The implementation checklist also covers encryption, export and deletion, disclosed third parties, PII-free logs, consent management, cookies, and backups.