The way an intervention is framed is important, because it often dictates the political strategy surrounding the use of force. In successful limited interventions such as Bosnia and Kosovo, the United States used force to drive the parties to the negotiating table. In Iraq and Afghanistan, the United States used force but never really had a clear vision for what a negotiated settlement would look like.
Another factor that has helped dictate whether past interventions grew into full-blown wars is the military technology available to the United States. The ability (separate from the willingness) to impose meaningful costs on a hostile regime or exact retribution for a violation of international norms from afar has increased as stand-off weaponry, intelligence and surveillance, and precision targeting capabilities have improved. This gives U.S. presidents the ability to stay out of more widespread interventions while still imposing (or appearing to impose) a cost on their opponents.
But the most crucial factor dictating whether limited interventions stay limited is a combination of the domestic political context in the United States and presidential willpower. Put simply, it matters whether the American people are willing to go to war. In 2001 and 2003, the American public wanted revenge for the attacks of 9/11, faced no other major national security threats, and had not experienced a major war in a generation. And it helped that for many on the right, the interventions were cast as part of a broader conflict against terrorism. The context is very different today, and the American people show little appetite for another ground war after more than a decade fighting in Iraq and Afghanistan.