<example> — when and how to use it in Claude prompts

Use one or more tags (usually inside ) for few-shot prompting.

When to use <example>

• Few-shot prompting — showing the model 1–5 demonstrations before the live input.
• Anchoring an unusual output format that's hard to describe in prose.

When not to use it

• Don't use a single <example> when zero-shot works — adding examples narrows the model's distribution and can hurt edge-case handling.
• Don't put real user data in examples; the model may pattern-match and leak details.

Minimal example

<example>
  <input>The food was cold and the waiter was rude.</input>
  <output>negative</output>
</example>

Full example

<instructions>
Classify customer reviews as positive, neutral, or negative. Return only the label.
</instructions>

<examples>
  <example>
    <input>Loved the new menu, will be back!</input>
    <output>positive</output>
  </example>
  <example>
    <input>Service was okay, food was okay.</input>
    <output>neutral</output>
  </example>
  <example>
    <input>Waited 40 minutes for cold food.</input>
    <output>negative</output>
  </example>
</examples>

<review>{{ user_review }}</review>

Common mistakes

• Wrapping the I/O directly with no <input>/<output> structure — Claude can usually figure it out, but the structured form is more reliable.
• Making examples that don't span the real distribution (e.g., all positive). Examples set Claude's prior; skew them and outputs skew with them.