Moderation Guidelines

This document outlines moderation policies, decision frameworks, and enforcement procedures.

Moderation Philosophy

Fair, consistent, and proportionate.

Protect the community from harmful content
Respect user privacy and freedom
Apply rules consistently across all users
Provide clear communication when taking action
Allow appeals and corrections
Document all decisions

Prohibited Content

Absolute Prohibitions (Immediate Removal)

Content that is automatically prohibited and removed on sight:

Malware & Exploits
- Executable files flagged by antivirus
- Exploit kits, weaponized PDFs
- Phishing kits, credential stealers
- Action: Remove immediately, no warning
- User notification: Standard (file was malware)
Child Exploitation
- Any CSAM (child sexual abuse material)
- Any sexualization of minors
- Action: Remove, ban user permanently, report to NCMEC
- User notification: None (direct law enforcement involvement)
Terroristic Content
- Bomb-making instructions
- Manifesto related to imminent violence
- Recruitment for terrorist organizations
- Action: Remove, ban, report to law enforcement
- User notification: None
Stolen Financial Information
- Credit card dumps
- Bank login databases
- PayPal/Venmo credentials
- Action: Remove, ban, investigate user
- User notification: We removed stolen financial data

Policy Violations (Conditional Enforcement)

Content that violates policies but may have legitimate context:

Copyright Infringement
- Published copyrighted works (books, movies, etc.)
- Software license cracks/keygens
- Assessment: Check for fair use (review, commentary, education)
- Action: If legitimate use → keep; Infringing → DMCA takedown
- User notification: Copyright claim explanation
Spam
- Identical files uploaded repeatedly
- Links to scams or malware
- Commercial promotion in public spaces
- Investigation: User behavior pattern analysis
- Action: Warning first → temporary suspension → permanent ban
- Escalation: If coordinated spam ring, involve law enforcement
Hate Speech
- Slurs or dehumanizing language targeting protected groups
- Assessment: Context (quote for discussion vs. endorsement)
- Action: First offense → warning; Repeat → temporary ban
- Appeals: Available via support email
Harassment & Threats
- Threats of violence against specific individuals
- Sustained harassment campaigns
- Doxxing (publishing private info)
- Assessment: Severity, targeting, pattern
- Action: Warn → ban from service if severe
- User protection: Offer to restrict profile visibility

Ban Severity Levels

Level	Duration	When to Use	Appeal
Warning	None	First minor violation	N/A
Temporary	7 days	Pattern of behavior	After suspension
Extended	30 days	Serious violation	After suspension
Permanent	Indefinite	Extreme/repeated violations	Email after 6 months

Moderation Process

Step 1: Report Received

When a report comes in:

Log into admin dashboard
Go to Reports section
Filter by status: open
Choose oldest report first (FIFO)
Click to open full report

Save the report ID for tracking.

Step 2: Initial Assessment

For each report, determine:

What is reported?
- File content
- User behavior
- Account activity
What is the violation?
- Which policy is broken?
- How severe?
Is there ambiguity?
- If yes, reach out to moderator supervisor
- If no, proceed
What action is appropriate?
- Warning
- Temporary ban
- Permanent ban
- Content removal
- Investigation

Step 3: Take Action

For content removal:

POST /api/admin/content/[fileId]
{
  "reason": "Copyright infringement",
  "notifyUser": true
}

For user warning:

Send email via /api/admin/emails/send
Subject: "Warning: Your account violated our terms"
Body: Explain violation and consequences

For user ban:

POST /api/admin/users/[userId]/ban
{
  "type": "temporary",
  "durationDays": 7,
  "reason": "Repeated copyright violations",
  "notifyUser": true
}

Step 4: Document Decision

Always document:

What was the report?
What did we find?
What action did we take?
Why did we take that action?

Update report status:

PATCH /api/admin/reports/[reportId]
{
  "status": "resolved",
  "action": "user_warned",
  "notes": "User warned about copyright violations. Explanation of policy sent via email."
}

Step 5: Communicate with User

When taking action, always notify the user unless it's a P1 offense (CSAM, malware, etc.)

Email Template - Warning:

Subject: Account Warning

Hi [Name],

We received reports about your activity on Emberly and found violations of our Terms of Service.

Violation: [Specific violation]
Details: [Explain what happened]

Actions you should take:
- [Stop doing X]
- [Don't do Y again]

If you believe this is a mistake, reply to this email within 7 days.

Next step: Another violation in 30 days will result in a temporary ban.

Thanks,
Emberly Moderation Team

Email Template - Ban:

Subject: Account Suspended

Hi [Name],

Your account has been temporarily suspended due to repeated violations of our Terms of Service.

Reason: [Explain clearly]
Duration: [X days]
Suspended at: [Date/time]
Will restore: [Date/time]

You can appeal this decision by replying to this email. Include:
1. Why you believe this was a mistake
2. How you'll avoid this in the future

Thanks,
Emberly Moderation Team

Appeal Process

Users can appeal any moderation decision except CSAM/malware/terrorism.

Appeal Workflow

User sends appeal email to support
Tag as [APPEAL] in subject
Assign to appeals specialist (usually a manager)
Appeals specialist:
- Reviews original evidence
- Considers appeal merit
- Checks for policy misapplication
- Makes decision (uphold, modify, reverse)
Send response to user

Timeline: Respond within 7 days

Escalation

When in doubt, escalate to your supervisor:

Ambiguous case? → Escalate
Multiple violations? → Escalate
Permanent ban? → Requires approval from manager
Security/legal issue? → Escalate to compliance team

Escalation email:

Subject: Moderation decision needed: [Report ID]

Context:
- What is reported
- Initial assessment
- Why you're unsure

Recommendation:
- What action do you think is appropriate
- Why

cc: supervisor

Common Scenarios

Scenario 1: Copyright Film

Report Content: Copyrighted movie file uploaded publicly

Investigation:

Is this for sale/commercial use? (Infringing)
Is it an excerpt for educational discussion? (Possibly fair use)
Is there copyright notice? (Supports infringement)

Decision: Check file size/details

Full movie (3GB) → Infringing, remove
Clip (100MB) with discussion → Examine context, likely keep
Promotional trailer → Likely fair use, keep

Action: If removing, DMCA notice to user. If keeping, document reason.

Scenario 2: Political Content

Report Content: File with "offensive political viewpoint"

Investigation:

Does it violate policy? (Policy bans illegal content, not opinions)
Is it hate speech? (Depends on language used)
Is reporter disagreeing with politics? (Not a violation)

Decision: Political speech is protected unless it meets hate speech threshold.

Action: Likely no action. Respond to reporter: "This doesn't violate our policy."

Scenario 3: Suspicious Account

Report Content: "User is violating TOS"

Investigation:

What is the actual violation?
Is there evidence?
Pattern of behavior?

Decision: Must see specific violation, not just suspicious behavior.

Action: If no clear violation, close report. If pattern detected, monitor account and investigate.

Tools You'll Use

Admin Dashboard

Reports page — See all reports, status, evidence
Users page — Search users, view history, take action
Audit logs — See all moderation actions taken
Flagged files — Review files marked suspicious

API Endpoints

Primary API calls you'll make:

# Get a report
curl -H "Authorization: Bearer ADMIN_KEY" \
  https://embrly.ca/api/admin/reports/report_123
 
# Update report status
curl -X PATCH -H "Authorization: Bearer ADMIN_KEY" \
  https://embrly.ca/api/admin/reports/report_123 \
  -d '{
    "status": "resolved",
    "action": "user_banned",
    "notes": "Pattern of spam"
  }'
 
# Ban a user
curl -X POST -H "Authorization: Bearer ADMIN_KEY" \
  https://embrly.ca/api/admin/users/user_456/ban \
  -d '{
    "type": "temporary",
    "durationDays": 7,
    "reason": "Copyright violations",
    "notifyUser": true
  }'
 
# Remove content
curl -X DELETE -H "Authorization: Bearer ADMIN_KEY" \
  https://embrly.ca/api/admin/content/file_789 \
  -d '{
    "reason": "Malware",
    "notifyUser": true
  }'

See Staff Admin API Reference for full documentation.

Metrics & Reporting

Weekly Metrics

Track and report:

Reports received: [N]
Reports resolved: [N]
Average resolution time: [X hours]
User bans: [N]
Appeals: [N] (and outcomes)

Monthly Report

Submit to ops lead:

Summary of moderation activity
Policy changes or clarifications needed
Escalations or edge cases
Team capacity/improvements

Special Cases

Selfie with ID Doxxing

User uploads photo with ID visible (can view home address, full name, etc.)

Action: Remove immediately, warn user "Don't share personal documents publicly"

TikTok/Instagram Download

User is sharing content from other platforms

Assessment:

Is it modified? (Stolen content)
Is it only a link? (Not infringing)
Is it their own content being re-shared? (OK)

Action: If stolen → Remove; If link/own content → OK

Bot Account Farming

Multiple accounts created by same person, spamming links

Investigation:

IP address check
Account creation pattern
Behavior similarity
File upload pattern

Action: Ban all linked accounts, potentially IP-ban if recurrent

FAQs About Moderation

Q: Can I ban someone for disagreeing with me? A: No. We don't ban people for opinions, only policy violations.

Q: What if someone appeals and they're right? A: Reverse the action, apologize sincerely, and document the error.

Q: Can I warn someone twice and then ban? A: Use progressive discipline, but severity of violation matters. One serious violation might lead to permanent ban.

Q: What if the report is fake? A: Close it and document. If same person submits false reports repeatedly, note the pattern.

Q: Am I allowed to ban myself to test the system? A: No. Use test accounts instead.

Support

Questions about moderation?

Ask in Slack: #emberly-moderation
Email supervisor
Review past decisions for precedent
Check internal wiki for new policy updates

On this page