Is there a security model for preventing injection attacks? I'm thinking of content-based injection attacks like those that try to break out of language parser

Is there a security model for preventing injection attacks? I'm thinking of content-based injection attacks like those that try to break out of language parser enclosures in HTML or Liquid or SQL or etc. Presumably the responsibility would largely fall on the component authors but I was curious if there were any language-level threat model or mitigation approach in mind here.