Meta-Prompting: AI That Writes AI Prompts

Use an AI to write, optimize, and debug your prompts. Meta-prompting is how top teams ship production prompts faster than humans alone could write them.

36 min · Reviewed 2026

The bootstrap move

Writing a great prompt is itself a task Claude is good at. Instead of crafting a perfect prompt by hand, describe the task to Claude and ask it to write the prompt for you. Anthropic ships an official prompt generator tool in the Console; under the hood, it's a very good meta-prompt.

A meta-prompt skeleton

You are an expert prompt engineer trained on the Anthropic prompt engineering guide.

My task:
{TASK_DESCRIPTION}

The model that will run the final prompt: Claude Sonnet 4.5.

Produce a production-ready prompt that:
1. Assigns a clear role.
2. States the task precisely.
3. Lists constraints as a numbered block.
4. Includes 2-3 few-shot examples (you may fabricate plausible ones).
5. Specifies the output format using XML tags.
6. Includes a chain-of-thought section if the task needs reasoning.

Output the prompt inside <prompt> tags and a short justification inside <rationale> tags.A meta-prompt that produces production prompts.

Run this with a task like 'I need to classify customer support emails as billing / technical / other, and route them.' Claude will produce a multi-section prompt complete with role, examples, and XML output — often better than what a busy engineer would hand-write in 10 minutes.

Prompt optimization loop

Collect 10 inputs with expected outputs (your eval set).
Run your current prompt on all 10. Record which fail.
Show Claude the prompt, the failing inputs, and the expected outputs.
Ask: 'What changes to this prompt would likely fix these failures without breaking the passing cases?'
Apply the suggested changes. Re-run the eval. Iterate.

I have this prompt: <prompt>...</prompt>

It passes these cases: <passes>...</passes>

It fails these cases: <failures>
  <case>
    <input>...</input>
    <expected>...</expected>
    <actual>...</actual>
  </case>
  ...
</failures>

Propose three distinct prompt revisions. For each:
- Explain the theory behind the change.
- Predict which failures it should fix.
- Flag any passing cases it might break.Forcing three revisions (not just one) surfaces tradeoffs.

Meta-prompt for evaluating prompts

You are a senior prompt engineering reviewer.

<prompt_to_review>
{PROMPT}
</prompt_to_review>

Evaluate it on:
1. Role clarity (1-5)
2. Instruction specificity (1-5)
3. Example quality (1-5)
4. Format precision (1-5)
5. Robustness to adversarial input (1-5)

For each score under 4, explain what's missing and give a concrete fix.A scoring rubric — turn prompt engineering into a measurable discipline.

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-prompting-meta-prompting-creators

What is the core idea behind "Meta-Prompting: AI That Writes AI Prompts"?
1. Use an AI to write, optimize, and debug your prompts. Meta-prompting is how top teams ship production prompts faster than humans alone could write them.
2. Paste 2 essays the teacher loved, then have AI copy that voice.
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
Which term best describes a foundational idea in "Meta-Prompting: AI That Writes AI Prompts"?
1. prompt generation
2. meta-prompting
3. prompt optimization
4. Paste 2 essays the teacher loved, then have AI copy that voice.
A learner studying Meta-Prompting: AI That Writes AI Prompts would need to understand which concept?
1. meta-prompting
2. prompt optimization
3. prompt generation
4. Paste 2 essays the teacher loved, then have AI copy that voice.
Which of these is directly relevant to Meta-Prompting: AI That Writes AI Prompts?
1. meta-prompting
2. prompt generation
3. Paste 2 essays the teacher loved, then have AI copy that voice.
4. prompt optimization
Which of the following is a key point about Meta-Prompting: AI That Writes AI Prompts?
1. Collect 10 inputs with expected outputs (your eval set).
2. Run your current prompt on all 10. Record which fail.
3. Show Claude the prompt, the failing inputs, and the expected outputs.
4. Ask: 'What changes to this prompt would likely fix these failures without breaking the passing cases…
Which of these does NOT belong in a discussion of Meta-Prompting: AI That Writes AI Prompts?
1. Show Claude the prompt, the failing inputs, and the expected outputs.
2. Collect 10 inputs with expected outputs (your eval set).
3. Run your current prompt on all 10. Record which fail.
4. Paste 2 essays the teacher loved, then have AI copy that voice.
What is the key insight about "Anthropic Console Prompt Generator" in the context of Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. First reply rarely nails it — that's expected.
3. The Anthropic Console includes a prompt generator and a prompt improver. Both are themselves high-quality meta-prompts.
4. Return only the requested artifact when constrained.
What is the key insight about "Keep the human in the loop" in the context of Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. First reply rarely nails it — that's expected.
3. Return only the requested artifact when constrained.
4. Meta-prompting accelerates — it doesn't replace judgment. Always read the generated prompt, test it on real edge cases, …
What is the recommended tip about "Practitioner tip" in the context of Meta-Prompting: AI That Writes AI Prompts?
1. Treat every prompt as a spec: role → context → task → format. Review your first output as a draft, not a final.
2. Paste 2 essays the teacher loved, then have AI copy that voice.
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
Which statement accurately describes an aspect of Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. Writing a great prompt is itself a task Claude is good at. Instead of crafting a perfect prompt by hand, describe the task to Claude and ask…
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
What does working with Meta-Prompting: AI That Writes AI Prompts typically involve?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. First reply rarely nails it — that's expected.
3. Run this with a task like 'I need to classify customer support emails as billing / technical / other, and route them.
4. Return only the requested artifact when constrained.
Which best describes the scope of "Meta-Prompting: AI That Writes AI Prompts"?
1. It is unrelated to prompting workflows
2. It applies only to the opposite beginner tier
3. It was deprecated in 2024 and no longer relevant
4. It focuses on Use an AI to write, optimize, and debug your prompts. Meta-prompting is how top teams ship productio
Which section heading best belongs in a lesson about Meta-Prompting: AI That Writes AI Prompts?
1. A meta-prompt skeleton
2. Paste 2 essays the teacher loved, then have AI copy that voice.
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
Which section heading best belongs in a lesson about Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. Prompt optimization loop
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
Which section heading best belongs in a lesson about Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. First reply rarely nails it — that's expected.
3. Meta-prompt for evaluating prompts
4. Return only the requested artifact when constrained.

← Back to interactive lesson

Tendril · Creators · Prompting

Meta-Prompting: AI That Writes AI Prompts

Use an AI to write, optimize, and debug your prompts. Meta-prompting is how top teams ship production prompts faster than humans alone could write them.

36 min · Reviewed 2026

The bootstrap move

A meta-prompt skeleton

You are an expert prompt engineer trained on the Anthropic prompt engineering guide.

My task:
{TASK_DESCRIPTION}

The model that will run the final prompt: Claude Sonnet 4.5.

Produce a production-ready prompt that:
1. Assigns a clear role.
2. States the task precisely.
3. Lists constraints as a numbered block.
4. Includes 2-3 few-shot examples (you may fabricate plausible ones).
5. Specifies the output format using XML tags.
6. Includes a chain-of-thought section if the task needs reasoning.

Output the prompt inside <prompt> tags and a short justification inside <rationale> tags.A meta-prompt that produces production prompts.

Prompt optimization loop

Collect 10 inputs with expected outputs (your eval set).
Run your current prompt on all 10. Record which fail.
Show Claude the prompt, the failing inputs, and the expected outputs.
Ask: 'What changes to this prompt would likely fix these failures without breaking the passing cases?'
Apply the suggested changes. Re-run the eval. Iterate.

I have this prompt: <prompt>...</prompt>

It passes these cases: <passes>...</passes>

It fails these cases: <failures>
  <case>
    <input>...</input>
    <expected>...</expected>
    <actual>...</actual>
  </case>
  ...
</failures>

Propose three distinct prompt revisions. For each:
- Explain the theory behind the change.
- Predict which failures it should fix.
- Flag any passing cases it might break.Forcing three revisions (not just one) surfaces tradeoffs.

Meta-prompt for evaluating prompts

You are a senior prompt engineering reviewer.

<prompt_to_review>
{PROMPT}
</prompt_to_review>

Evaluate it on:
1. Role clarity (1-5)
2. Instruction specificity (1-5)
3. Example quality (1-5)
4. Format precision (1-5)
5. Robustness to adversarial input (1-5)

For each score under 4, explain what's missing and give a concrete fix.A scoring rubric — turn prompt engineering into a measurable discipline.

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-prompting-meta-prompting-creators

What is the core idea behind "Meta-Prompting: AI That Writes AI Prompts"?
1. Use an AI to write, optimize, and debug your prompts. Meta-prompting is how top teams ship production prompts faster than humans alone could write them.
2. Paste 2 essays the teacher loved, then have AI copy that voice.
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
Which term best describes a foundational idea in "Meta-Prompting: AI That Writes AI Prompts"?
1. prompt generation
2. meta-prompting
3. prompt optimization
4. Paste 2 essays the teacher loved, then have AI copy that voice.
A learner studying Meta-Prompting: AI That Writes AI Prompts would need to understand which concept?
1. meta-prompting
2. prompt optimization
3. prompt generation
4. Paste 2 essays the teacher loved, then have AI copy that voice.
Which of these is directly relevant to Meta-Prompting: AI That Writes AI Prompts?
1. meta-prompting
2. prompt generation
3. Paste 2 essays the teacher loved, then have AI copy that voice.
4. prompt optimization
Which of the following is a key point about Meta-Prompting: AI That Writes AI Prompts?
1. Collect 10 inputs with expected outputs (your eval set).
2. Run your current prompt on all 10. Record which fail.
3. Show Claude the prompt, the failing inputs, and the expected outputs.
4. Ask: 'What changes to this prompt would likely fix these failures without breaking the passing cases…
Which of these does NOT belong in a discussion of Meta-Prompting: AI That Writes AI Prompts?
1. Show Claude the prompt, the failing inputs, and the expected outputs.
2. Collect 10 inputs with expected outputs (your eval set).
3. Run your current prompt on all 10. Record which fail.
4. Paste 2 essays the teacher loved, then have AI copy that voice.
What is the key insight about "Anthropic Console Prompt Generator" in the context of Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. First reply rarely nails it — that's expected.
3. The Anthropic Console includes a prompt generator and a prompt improver. Both are themselves high-quality meta-prompts.
4. Return only the requested artifact when constrained.
What is the key insight about "Keep the human in the loop" in the context of Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. First reply rarely nails it — that's expected.
3. Return only the requested artifact when constrained.
4. Meta-prompting accelerates — it doesn't replace judgment. Always read the generated prompt, test it on real edge cases, …
What is the recommended tip about "Practitioner tip" in the context of Meta-Prompting: AI That Writes AI Prompts?
1. Treat every prompt as a spec: role → context → task → format. Review your first output as a draft, not a final.
2. Paste 2 essays the teacher loved, then have AI copy that voice.
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
Which statement accurately describes an aspect of Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. Writing a great prompt is itself a task Claude is good at. Instead of crafting a perfect prompt by hand, describe the task to Claude and ask…
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
What does working with Meta-Prompting: AI That Writes AI Prompts typically involve?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. First reply rarely nails it — that's expected.
3. Run this with a task like 'I need to classify customer support emails as billing / technical / other, and route them.
4. Return only the requested artifact when constrained.
Which best describes the scope of "Meta-Prompting: AI That Writes AI Prompts"?
1. It is unrelated to prompting workflows
2. It applies only to the opposite beginner tier
3. It was deprecated in 2024 and no longer relevant
4. It focuses on Use an AI to write, optimize, and debug your prompts. Meta-prompting is how top teams ship productio
Which section heading best belongs in a lesson about Meta-Prompting: AI That Writes AI Prompts?
1. A meta-prompt skeleton
2. Paste 2 essays the teacher loved, then have AI copy that voice.
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
Which section heading best belongs in a lesson about Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. Prompt optimization loop
3. First reply rarely nails it — that's expected.
4. Return only the requested artifact when constrained.
Which section heading best belongs in a lesson about Meta-Prompting: AI That Writes AI Prompts?
1. Paste 2 essays the teacher loved, then have AI copy that voice.
2. First reply rarely nails it — that's expected.
3. Meta-prompt for evaluating prompts
4. Return only the requested artifact when constrained.

← Back to interactive lesson