Prompt (rr-iter2-013): Explain how `skill_advisor.py` computes uncertainty.
Expected Gate 3: false

Classifier output:
{
  "triggersGate3": false,
  "reason": "no_match",
  "matched": [],
  "readOnlyMatched": [
    {
      "pattern": "explain",
      "kind": "token",
      "category": "read_only"
    }
  ]
}