Input guardrails
Input guardrails overview
Input guardrails give administrators and agent builders control over what users ask . The rules apply however users reach , including chat, integrations, and autonomous triggered agents.
Create and train an input guardrail
Create an input guardrail to enforce a content policy on the messages users send to . This article covers writing a rule, training a classifier, tuning the results, and testing the guardrail.
Write effective input guardrail rules
The rule text you write determines how accurately an input guardrail blocks the right messages. A specific, focused rule trains a reliable classifier.
Activate an input guardrail for your instance
Activate an input guardrail to enforce one content policy for your instance of . The instance-level input guardrail applies to the requests handles directly.
Assign an input guardrail to a specialized agent
Assign an input guardrail to an individual specialized agent when separate agents handle distinct conversations and need their own content policies. A per-agent guardrail enforces its policy only on messages sent to that agent.
Restore a previous guardrail version
Every change to a guardrail creates a new immutable version in . Browse the version history to review previous versions and restore any of them.
Input guardrails FAQ
Find answers to common questions about configuring, training, and troubleshooting input guardrails in . For step-by-step instructions, see Create and train an input guardrail .
Manage input guardrails
Manage your input guardrails from the Input Guardrails page after you create them in . Update a guardrail to change its rule, name, or block message, or delete a guardrail your team no longer needs.