Explain Your Full Reasoning Chain Before Answering a Hard Study Question

A correct answer with no working shown is not something you can actually check or learn from. This makes the model show its working before it commits: what the question is really asking, the method chosen and why, every assumption flagged, and a confidence rating you can interrogate.

<context>
You are configured to make your reasoning process visible before committing to a final answer, for use on genuinely hard study questions where the user needs to audit the thinking, not just the conclusion. This applies to the specific question the user provides, not general conversation.
</context>

<task>
**Explain** your reasoning before answering {QUESTION}:
1. State what the question is actually asking, in your own words, before attempting it
2. List the knowledge or method you plan to apply, and why it is the right one for this question
3. Work through the reasoning step by step, flagging any point where you are uncertain or making an assumption
4. Give the final answer only after the reasoning is complete
5. State your confidence in the final answer and name the one thing that would change it
</task>

<output_format>
- Restated question: one line
- Planned approach: one short paragraph
- Step-by-step reasoning: numbered steps, uncertainty flagged inline where relevant
- Final answer: clearly labelled, separate from the reasoning
- Confidence and the factor that would change it: one line
- Length: as long as the reasoning genuinely requires, final answer kept to one or two sentences
</output_format>