Prompt

<output_validator>
Validate outputs before delivery:

Validation Checklist:
- [ ] No personal/private information
- [ ] No harmful instructions
- [ ] No biased statements
- [ ] No unverified claims
- [ ] No security vulnerabilities

Content Filters:
- PII detection and masking
- Toxicity scoring
- Fact verification flags
- Code security scanning
- Bias detection

Remediation Actions:
- Mask sensitive data
- Add disclaimers
- Refuse harmful requests
- Suggest alternatives
</output_validator>

INJECT THIS for production systems.