GDPR Procedures

Overview

SanMarcSoft services operate under EU jurisdiction (infrastructure in Scaleway fr-par). All services must comply with GDPR requirements for data subject rights. This runbook covers handling data subject requests (DSRs) for access, erasure, and portability.

Data Subject Rights

Service-by-Service Data Inventory

Verifieddit (verifieddit.com)

Data stored:

• Clerk user profile (email, name, auth provider)
• D1 badges table (URLs, domains, timestamps)
• D1 badge_images (image URLs, AI scores)
• D1 users table (clerk_id, email, plan)

Data locations:

• Clerk (US-hosted, EU data processing agreement)
• Cloudflare D1 (edge, geo-distributed)
• Scaleway Container Registry (fr-par, EU sovereign)

Phenom Drop

Data stored:

• Submitted media files (S3)
• Firestore drop payloads (email, consent records)
• C2PA signed media

Data locations:

• AWS S3 (us-east-1)
• Google Firestore
• AWS App Runner (ephemeral)

Badges Worker

Data stored:

• Badge records (D1)
• Cached AI detection results

Data locations:

• Cloudflare D1 (edge)
• Cloudflare KV (edge, if caching enabled)

Procedure: Right of Access (Art. 15)

Step 1: Identify the Data Subject

Verify the requester’s identity. Require government-issued ID or verification via their registered email.

Step 2: Collect Data from Each Service

From Clerk

1
2
3
4
5
6

CLERK_SECRET=$(pass verifieddit/clerk/secret-key)

# Search by email
curl -s -H "Authorization: Bearer ${CLERK_SECRET}" \
  "https://api.clerk.com/v1/users?email_address=user@example.com" | \
  jq '.data[] | {id, email: .email_addresses, first_name, last_name, created_at, last_sign_in_at}'

From D1

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

# Get user record
npx wrangler d1 execute verifieddit-badges \
  --command "SELECT * FROM users WHERE email = 'user@example.com'" --json

# Get associated badges
CLERK_ID="<clerk-id-from-above>"
npx wrangler d1 execute verifieddit-badges \
  --command "SELECT * FROM badges WHERE user_id = '${CLERK_ID}'" --json

# Get badge images
npx wrangler d1 execute verifieddit-badges \
  --command "SELECT bi.* FROM badge_images bi JOIN badges b ON bi.badge_id = b.id WHERE b.user_id = '${CLERK_ID}'" --json

From Phenom Drop (Firestore)

Query Firestore for records matching the user’s email:

1
2

# This requires Firebase Admin SDK access
# Query: firestore.collection('drops').where('submitterEmail', '==', 'user@example.com')

Step 3: Compile and Send

1. Compile all data into a structured format (JSON or PDF)
2. Send to the requester via their verified email
3. Log the access request and response date

Procedure: Right to Erasure (Art. 17)

Step 1: Verify Identity

Same as Right of Access.

Step 2: Delete from Each Service

Delete from Clerk

1
2
3
4
5
6

CLERK_SECRET=$(pass verifieddit/clerk/secret-key)
USER_ID="<clerk-user-id>"

curl -s -X DELETE \
  -H "Authorization: Bearer ${CLERK_SECRET}" \
  "https://api.clerk.com/v1/users/${USER_ID}"

Delete from D1

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13

CLERK_ID="<clerk-id>"

# Delete badge images first (foreign key)
npx wrangler d1 execute verifieddit-badges \
  --command "DELETE FROM badge_images WHERE badge_id IN (SELECT id FROM badges WHERE user_id = '${CLERK_ID}')"

# Delete badges
npx wrangler d1 execute verifieddit-badges \
  --command "DELETE FROM badges WHERE user_id = '${CLERK_ID}'"

# Delete user record
npx wrangler d1 execute verifieddit-badges \
  --command "DELETE FROM users WHERE clerk_id = '${CLERK_ID}'"

Delete from Phenom Drop

1. Delete S3 media files:

1

aws s3 rm s3://phenom-drop-media/drops/<drop-id>/ --recursive

2. Delete Firestore records:

1
2

# Via Firebase Admin SDK or console
# Delete all documents where submitterEmail matches

Step 3: Confirm Deletion

1. Verify deletion by re-querying each service
2. Send confirmation to the data subject
3. Log the erasure request and completion date

Exceptions to Erasure

Data may be retained if required for:

• Legal compliance (e.g., tax records)
• C2PA provenance chain integrity (content credentials are permanent by design – inform the user)
• Ongoing legal proceedings

Procedure: Right to Data Portability (Art. 20)

Step 1: Export Data

Follow the same collection steps as Right of Access, but export in a machine-readable format (JSON).

Step 2: Package and Deliver

1
2
3
4

# Create export package
mkdir -p export-${USER_EMAIL}
# Save each service's data as separate JSON files
# Zip and send via secure channel

Request Tracking

Log all DSR requests with:

• Request date
• Data subject identifier
• Request type (access/erasure/portability)
• Services affected
• Completion date
• Notes

Response Timeline

• Acknowledge receipt within 3 business days
• Complete the request within 30 calendar days
• If complex, extend to 60 days with notification to the data subject

Troubleshooting

• Cannot find user data: Search by all possible identifiers (email, Clerk ID, Firestore document ID).
• Clerk user already deleted: Data may still exist in D1 and S3. Clean up orphaned records.
• S3 objects not deletable: Check bucket versioning. If enabled, objects are soft-deleted and must be permanently removed.