ClaudeCraft

Interactive Playground

The difference between a "good enough" LLM response and a production-grade one usually comes down to prompt structure. Expand the examples below to compare approach and output.

Refactoring a Messy Function

See how adding structure transforms a vague refactoring request into clean, targeted output.

text
Refactor this function:

function processData(d) {
  var r = [];
  for (var i = 0; i < d.length; i++) {
    if (d[i].status == 'active' && d[i].age > 18) {
      r.push({ name: d[i].firstName + ' ' + d[i].lastName, age: d[i].age });
    }
  }
  return r;
}

Ready to structure your own prompts?

Use our XML Tag Generator to wrap your tasks, context, and examples in Claude-native XML boundaries.

Open XML Generator