Guardrail
Guardrail — on this site: a decision contract that screens untrusted text (prompt, completion, or tool call) with Jev Noul/Score/Choice questions, then lets application code allow, review, or block.
It is not TypeSafe’s product name for a firewall. Official docs describe the pattern under the LLM guardrails cookbook and “Universal Verification” on the use-case map. jev-1.13 is not hardened against adversarial state by default.
See also: guardrails workflow, policy checks, LLM guardrails.
Unofficial definition. Confirm cookbooks at docs.typesafe.ai.
Sources
Public TypeSafe or adjacent documentation only. No private claims.