Avisa AíAvisa Aí
Security

Security and privacy at every layer

Avisa Aí’s protections are concrete technical decisions enforced by the server, not generic promises. This page explains what happens to accounts, access and data.

Accounts and sessions

  • Two-step verification

    Optional, via authenticator app (with single-use backup codes) or a code sent by email. The code expires within minutes and allows only a few attempts.

  • Self-renewing sessions

    Access is short-lived and renewed with a rotating token. If an old token is reused (a sign of theft), all of that person’s sessions are ended.

  • Protected passwords and tokens

    Passwords are stored with a modern, brute-force-resistant hashing algorithm, and tokens only as hashes. On the web dashboard, the session lives in a cookie that page scripts cannot read.

  • Confirmed email to write

    Accounts created with email and password can only report, comment or vote after confirming the address. Sign-ins with external accounts arrive already verified.

Access control

Nobody sees or changes more than they need to, and no author approves their own report.

  • Access by role, city and category

    An officer can only act on the categories and city they are responsible for; a manager, only in their city. These rules are checked on the server on every action.

  • Protected by default

    Every server route requires login unless explicitly marked public: forgetting a guard does not open a hole.

  • On-site presence

    Approving, starting, completing and validating a report requires being within 100 m of the problem, which makes remote fraud harder.

  • Auditable history

    Every status change records who acted, when, with what justification and evidence, in a single transaction with the report itself.

Abuse limits

Limits per kind of action, stricter where a brute-force attempt would make the most sense.

  • Login
  • Two-step verification code
  • Sign-up and password recovery
  • Comments
  • Votes
  • Report creation

Content and personal data

  • Verified files

    Each photo or video must belong to whoever attached it, respect the size limit and actually contain the declared type. A file disguised as an image is refused and deleted.

  • Flagging, blocking and censoring

    Inappropriate content can be flagged, users can be blocked and offensive language is censored automatically in reports and comments.

  • CPF never exposed

    The system never tells unauthenticated users whether a CPF (Brazilian taxpayer ID) is already registered, and validates the check digits before any lookup.

  • Account deletion

    People can delete their account at any time, with a 30-day grace period. After that, personal data is removed and reports remain in the public history without identification.

  • Secrets out of the code

    Database passwords, keys and credentials live in a secrets vault, never in the repository. Services run without administrator privileges.

  • Continuous checks

    On every release, the code goes through static security analysis and secret scanning, the server image through vulnerability scanning and the test environment through dynamic scanning.

Still have questions?

See the most common questions about the app and the dashboard.