Crivora · Information Security Policy
Security at Crivora
We take the protection of Crivora's data and integrations seriously.
- Policy owner
- Crivora
- Effective date
- Last reviewed
- Review cadence
- at least annually and when material changes occur
- Version
- 1.0
- Contact
- contato@usecrivora.com
- URL
- https://usecrivora.com/security/en
1. Security governance
This policy establishes the information security program of Crivora (https://usecrivora.com): the public website, the signed-in app, the API routes, file storage and the integrations with third parties (payments, AI image generation and TikTok Shop).
Crivora's technical owner owns this policy: approves policies, defines controls, oversees their application and is accountable for incidents. Crivora is an early-stage SaaS; controls are proportional to the current operation and are reviewed at least annually and whenever the architecture or the integrations change.
Crivora does not currently hold security certifications or independent security audits.
- Access Control Policy: /security/access-control/en
- Security Baseline: /security/baseline/en
- Data Classification: /security/data-classification/en
- Incident Response: /security/incident-response/en
- Vulnerability Management: /security/vulnerability-management/en
- Personal Data Protection: /security/data-protection/en
- Subprocessors: /subprocessors/en
2. Access control
- The app area requires sign-in; without a valid session, access is redirected to the sign-in page.
- On the server, every data and file query is scoped to the owning account. A user cannot read another user's projects, images or generations.
- Internal and administrative routes are restricted to an authorized user configured on the server or to a server-side secret, and are unavailable without that configuration.
- Access to hosting, the database and provider dashboards is restricted to the technical owner. Details in /security/access-control/en.
3. Least privilege
- Each integration uses only the permissions its function requires (for example, read access to TikTok Shop rankings).
- Each user can access only their own data.
- Internal functions are restricted to those who need them and are reviewed when no longer needed.
4. Authentication
- Passwords are never stored in plain text: we only store a bcrypt hash.
- The session is kept in a cookie with the HttpOnly, Secure and SameSite attributes, managed by the authentication framework.
- Critical provider accounts (hosting, source code, email, payments and integrations) must use multi-factor authentication, as defined in /security/baseline/en.
5. Secret management
- API keys, tokens and service secrets are kept in server environment variables, outside the source code repository.
- Secrets are not sent to the browser or included in client-side code, and are never displayed in the interface.
- Webhook and internal-route secrets are validated on the server using constant-time comparison.
- If exposure is suspected, the secret is revoked and replaced (see /security/incident-response/en).
6. Encryption
Application controls (implemented by Crivora):
- In transit: public traffic to the website and the app uses HTTPS (TLS), with HSTS enabled.
- TikTok Shop OAuth tokens: encrypted with AES-256-GCM before being stored, with a key kept only on the server and separate from the database.
- Passwords: protected with bcrypt hashing (non-reversible).
- Service secrets: kept only in the server environment.
Infrastructure controls (hosting provider): Crivora's database and files run on Hostinger hosting services, whose Data Processing Addendum (https://www.hostinger.com/legal/dpa) lists, among the technical measures Hostinger maintains, encryption of personal and sensitive data at rest (databases, shared directories) and encryption of communications via strong cryptographic protocols.
In short: Crivora uses application-level encryption for selected restricted credentials and relies on documented infrastructure safeguards from its hosting provider for encryption of personal and sensitive data at rest. Crivora does not directly manage the provider's disk encryption.
7. Application security
- Inputs are validated on the server; data coming from the browser is never treated as trusted.
- Routes that change data verify the request origin (CSRF protection), in addition to the built-in protection of Server Actions.
- Uploads: the format is detected from the file content, size and dimension limits apply, and images are re-encoded with metadata (such as location) removed.
- HTTP security headers: X-Content-Type-Options, Referrer-Policy, Permissions-Policy, protection against being framed by other sites, and HSTS.
- Errors are handled in a controlled way, without exposing internal details to users.
- TikTok Shop integration: official OAuth authorization, random single-use short-lived state parameter, server-side signed requests; app pages read stored data and do not call the TikTok API while users browse.
- Payments: checkout hosted by Asaas; Crivora does not receive card data. Paid access is granted only by a webhook validated on the server, never by returning from checkout alone.
8. Data protection
- Data is classified according to /security/data-classification/en, with requirements per level.
- Uploaded and generated images are private and delivered through authenticated routes that check the file owner.
- Temporary uploads not used in projects are removed by periodic cleanup (after about 24 hours).
- Personal data protection principles: /security/data-protection/en. Processing details: /privacy/en.
9. Logging
- Operational errors may be logged for diagnosing and fixing problems.
- Sensitive information is not deliberately written to logs. Tokens, secrets, passwords and credentials are not included in application logs.
- Logs are reviewed by the technical owner when needed to investigate failures or suspected abuse.
10. Vulnerability management
Vulnerabilities in dependencies, the framework and the code are reviewed, prioritized by severity and risk, remediated and verified. The process is described in /security/vulnerability-management/en.
11. Incident response
Incidents follow the process detect → triage → contain → investigate → eradicate → recover → notify when applicable → review. Roles, communication and data breach notification: /security/incident-response/en.
12. Third parties
We use only the providers needed to operate the service, each with access limited to its function. The current list is at /subprocessors/en.
13. Business continuity
- The application, database and files run on a managed hosting provider.
- Source code is version-controlled in a private Git repository, which allows the application to be restored or redeployed.
- Data recovery relies on the hosting provider's backup capabilities. We do not currently publish formal recovery time objectives.
14. Responsible disclosure
If you believe you found a vulnerability, email contato@usecrivora.com with a description, steps to reproduce and the impact you observed.
- Do not access, modify or delete other people's data.
- Do not degrade the service (no denial-of-service testing, spam or social engineering).
- Allow a reasonable time for a fix before any public disclosure.
We review every good-faith report. We do not currently offer a bug bounty program.
15. Contact
Email: contato@usecrivora.com
Standard contact file: /.well-known/security.txt
Trust Center: /trust
16. Last updated
Version 1.0, effective October 9, 2026.
Crivora · Security at Crivora · https://usecrivora.com/security/en · contato@usecrivora.com