Skip to main content

​
Overview

Percify enforces safety through proactive account bans, visibility controls, and standardized API denial responses. This page summarizes mechanisms derived from the internal Ban System Guide.

​
Ban System

FieldPurpose
bannedBoolean flag blocking feature use
banned_atTimestamp of ban action
banned_reasonHuman-readable rationale
Utility functions:
  • checkUserBanned(userId) returns status
  • requireNotBanned(userId) throws if banned
  • createBannedUserResponse() returns 403 standardized JSON

​
Integration Pattern (Route Snippet)

const session = await auth.api.getSession({ headers: request.headers });
if (!session?.user) return NextResponse.json({ error: 'Unauthorized' }, { status: 401 });
try { await requireNotBanned(session.user.id); } 
catch { return createBannedUserResponse(); }

​
Protected Endpoints (High Priority)

  • Avatar generation & management
  • Video & audio creation routes
  • Profile update (handle, bio, avatar image)

​
Visibility Controls

StateIndexedPublic FeedAPI PublicOwner Access
DraftNoNoAuth onlyYes
PrivateYes (limited)NoAuth onlyYes
PublishedYesYesPublicYes

​
Content Moderation (Recommended)

LayerStrategy
Prompt FilteringBlock disallowed terms pre-generation
Post-Gen ReviewFlag suspicious outputs for manual audit
Rate LimitsThrottle abuse patterns
Ban EscalationAutomated triggers (credit exploit, spam)

​
Exploit Mitigations

Internal fixes cover:
  • Credit real-time synchronization to stop race conditions
  • Hardening against duplicate API calls (idempotency)
  • Centralized cost logic removal of scattered constants

​
Standard Error Responses

StatusCaseBody
401Missing auth{ "error": "Unauthorized" }
403Banned user{ "error": "Access denied", "reason": "banned" }
429Rate limit{ "error": "Too Many Requests" }
400Validation{ "error": "Invalid input" }

​
Logging & Audit

Log structure suggestion: 
{
  "type": "ban_check",
  "userId": "usr_123",
  "result": "blocked",
  "route": "/api/images/generate",
  "timestamp": "2025-11-24T12:45:00Z"
}

​
User Data Privacy (High Level)

  • Minimal PII stored (email, optional display name)
  • Media assets segregated by userId
  • Token-based API access (no password sharing)
  1. Detect anomaly (logs / alerts)
  2. Temporarily ban account (script)
  3. Archive related assets
  4. Review root cause
  5. Restore or escalate permanent ban
  • [/percify/credits]
  • [/percify/performance]
  • [/percify/faq]
Performance safeguards next: [/percify/performance].