Guardrails
Real-time firewall rules that detect and redact secrets, PII, and toxic content in inbound prompts
SUPERWISE® Sentinel operates natively at the edge, parsing every inbound prompt in real time before it reaches the upstream provider. If a prompt violates your organizational compliance policy, the Sentinel intervenes immediately — redacting the offending content so sensitive data never leaves your perimeter and malicious inputs never reach upstream models.
Guardrails run on inbound prompts. For each one, you choose whether to remediate — replace matched content with your placeholder text and forward the request — or block the request outright. See Manage guardrails.
Available guardrails
- Secret & credential detection — catch API keys, cloud credentials, and private tokens before they reach the model.
- PII redaction — catch emails, phone numbers, government and financial IDs, and other personal identifiers.
- Rudeness & toxicity — catch profanity, hostile language, and harassment.
Each Sentinel runs its own policy — see Manage guardrails to add, remove, configure, and order guardrails on a specific gateway.
Updated 21 days ago