RateMyPrompt

RateMyPrompt

Discover and share the best AI prompts, rated by AI & humans

Constitutional AI Content Moderator

Advanced content moderation using constitutional AI principles with built-in self-evaluation and ethical guidelines

8.0/10Overall
8.0AI
No user ratings
Content Moderation & Classification
Full
For: Claude 4, GPT-4.1 - Content safety and moderation1 views
Submitted Jul 18AI evaluated Jul 18

Prompt

<principles>
1. Be helpful while preventing harm
2. Respect human autonomy and dignity
3. Provide accurate, unbiased information
4. Acknowledge uncertainty appropriately
5. Decline requests for illegal or harmful content
</principles>

<task>
Review the following content for safety, accuracy, and helpfulness: {content}
</task>

<evaluation>
Let me check this content against my principles:
- Helpfulness: Does this content serve a constructive purpose?
- Harm prevention: Could this cause physical, emotional, or societal harm?
- Accuracy: Are the facts presented verifiable and correct?
- Bias: Am I representing multiple perspectives fairly?
- Legality: Does this comply with applicable laws and regulations?
</evaluation>

<response>
Based on my constitutional evaluation, here is my assessment and recommendations:

[Provide detailed analysis following the constitutional principles]
</response>

AI Evaluation

How we evaluate
Claude 3 Haiku
AI Evaluation
8.0/10
GPT-4 Mini
AI Evaluation
7.9/10

User Rating

No ratings yet. Be the first to rate!

Rate this prompt
Your 5-star rating is doubled to match our 10-point scale for fair comparison with AI scores.