I've written before about how we have little to fear and much to gain from state-called conventions to amend the Constitution. Even setting aside the legal and historical points about how such conventions can be limited and are the ultimate guarantors of state and popular sovereignty, the "runaway convention" idea is a red herring for one practical reason: Any proposed constitutional amendment emerging from the convention requires ratification by 38 states. That's at least 75 houses (counting Nebraska's unicameralism)! There is no way that an amendment constitutionalizing socialism (or whatever) would be ratified. So I sleep easy.
But to add further scholarly guidance to these arguments and procedures, see Rob Natelson's final paper in the Goldwater Institute's three-part series on the subject. From the executive summary:
This report provides crucial practical drafting guidance for exercising the states’ constitutional authority. In essence, it recommends that state legislators draft their Article V applications and delegate commissions with an eye to targeting specific subject matters, while still giving state delegates a meaningful level of deliberative independence to ensure that the amendments convention can serve its consensus-building and problem-solving purpose. The key is to regard an amendments convention as a modern-day “task force”—a representative body that is limited to a specific agenda but expected to exercise judgment on accomplishing that agenda.
For the previous two papers, and other materials regarding amendment conventions, see Goldwater's invaluable Artivle V resource page.