Collected sources and patterns will appear here. Add from search, explore, or the patterns library.
RawText -> ValidatedText
Evaluate incoming inputs or outgoing model responses against configurable safety, schema, or compliance rules.
Problem it solves
Unchecked agent inputs or generations can violate safety parameters, leak data, or break schema requirements.
Consumes
Emits
The real projects this mechanism was found in. Attribution is the point — this is how the best teams actually do it.