On Making Indirect Proofs Direct