Creator Profiles

​

Overview

​

Purpose

• Store encrypted platform credentials using AES-256-GCM server-side encryption
• Track scraping status and last successful sync timestamps
• Manage platform-specific configuration and authentication state
• Link platform accounts to user personas for multi-platform management

​

Key Fields

​

Credential Encryption

// Encryption format (server-side only)
{
  enc: "v1:iv:tag:ciphertext"
}

• All encryption happens server-side only
• Encryption key stored in CREDENTIALS_ENC_KEY_B64 environment variable
• No encryption keys are ever sent to the browser
• See server/utils/credentialsCrypto.js for implementation details

​

Example Queries

​

List All Creator Profiles

// Using directus MCP server
await use_mcp_tool({
  server_name: "directus",
  tool_name: "read-items",
  arguments: {
    collection: "creator_profiles",
    fields: ["id", "platform", "username", "scrape_status", "last_scraped_at"],
    filter: {
      is_active: { _eq: true }
    }
  }
});

​

Get Single Profile

await use_mcp_tool({
  server_name: "directus",
  tool_name: "read-item",
  arguments: {
    collection: "creator_profiles",
    id: "profile-uuid-here",
    fields: ["*"]
  }
});

​

Update Scrape Status

await use_mcp_tool({
  server_name: "directus",
  tool_name: "update-item",
  arguments: {
    collection: "creator_profiles",
    id: "profile-uuid-here",
    data: {
      scrape_status: "success",
      last_scraped_at: new Date().toISOString()
    }
  }
});

​

Search Profiles by Platform

await use_mcp_tool({
  server_name: "directus",
  tool_name: "search-items",
  arguments: {
    collection: "creator_profiles",
    query: "onlyfans",
    fields: ["id", "platform", "username", "is_active"]
  }
});

​

Related Collections

• platform_sessions - Encrypted browser cookies from extension for authentication
• media_jobs - Scraping jobs triggered for this profile
• scraped_media - Content scraped from this profile
• hitl_sessions - Human-in-the-loop login requests when scraping requires manual intervention

​

Workflow Integration

​

Profile Scraping Flow

1. User connects platform via /app/platforms
2. System checks for existing cookies in platform_sessions
3. If cookies exist: Create media_jobs entry with type scrape_profile
4. If no cookies: Create hitl_sessions entry and show yellow dashboard alert
5. Media worker processes job using Stagehand browser automation
6. Results stored in scraped_media, profile’s scrape_status updated

​

Credential Management

• Encryption endpoint: POST /api/credentials/encrypt
• Decryption endpoint: POST /api/credentials/decrypt
• Implementation: server/utils/credentialsCrypto.js
• Format: Legacy string support + new object format {enc: "v1:..."}

​

Best Practices

1. Never decrypt credentials in browser - All encryption operations are server-side only
2. Check platform_sessions before scraping - Avoids unnecessary HITL requests
3. Monitor scrape_status - Poll for job completion rather than blocking
4. Respect rate limits - Use last_scraped_at to prevent over-scraping
5. Validate platform field - Ensure platform is in supported list before creating profile

​

See Also

• Media Collections - Scraped content and scheduled posts
• Jobs Collections - Background job tracking
• Directus MCP Tools - Full tool reference