Prompt Optimisation with OpenAIs Meta-Prompt

OpenAI has just dropped a Meta-Prompt. It's designed to make crafting and fine-tuning prompts for language models a whole lot easier. The promise? To save you time and effort while boosting the quality of your AI-driven instructions. Sounds like a win-win, right?

Here’s the deal: the Meta-Prompt is baked right into OpenAI’s Playground, where it acts like a smart assistant for your language model. Whether you’re starting from scratch or improving an existing prompt, the Meta-Prompt steps in with best practices and real-world insights to guide the model in delivering what you want, faster and smarter.

Less Guesswork, More Structure

Creating a good prompt can feel like throwing spaghetti at the wall to see what sticks. OpenAI wants to change that by giving you a clear recipe for success. The Meta-Prompt starts with a straight-to-the-point task description, followed by extra details and optional sections to lay out specific steps, output format, examples, and more.

Here’s what it brings to the table:

1. Task Understanding: The model locks in on the main goal and expected output, ensuring it knows what you’re after.
2. Minimal Tweaks: If you’re working with an existing prompt, the Meta-Prompt suggests only simple improvements, focusing on clarity without overhauling the whole thing.
3. Reasoning Over Results: It prioritizes logical steps before jumping to conclusions, encouraging a deeper level of reasoning.
4. Examples Done Right: High-quality, relevant examples are included where needed, using placeholders for trickier bits.
5. Clarity and Brevity: The language is sharp, specific, and free of unnecessary fluff.
6. Markdown Magic: For better readability, it makes use of markdown—but skips code blocks unless explicitly needed.
7. Preserving User Input: Any detailed guidelines or examples you provide? It keeps them intact as much as possible.
8. Defined Output Format: It doesn’t leave you guessing how the results will be structured—it spells out the format clearly.

How It Works

Want to see the Meta-Prompt in action? Here’s a peek at the code that powers it:

from openai import OpenAI

client = OpenAI()

META_PROMPT = """
Given a task description or existing prompt, produce a detailed system prompt to guide a language model in completing the task effectively.

# Guidelines

- Understand the Task: Grasp the main objective, goals, requirements, constraints, and expected output.
- Minimal Changes: If an existing prompt is provided, improve it only if it's simple. For complex prompts, enhance clarity and add missing elements without altering the original structure.
- Reasoning Before Conclusions**: Encourage reasoning steps before any conclusions are reached. ATTENTION! If the user provides examples where the reasoning happens afterward, REVERSE the order! NEVER START EXAMPLES WITH CONCLUSIONS!
- Reasoning Order: Call out reasoning portions of the prompt and conclusion parts (specific fields by name). For each, determine the ORDER in which this is done, and whether it needs to be reversed.
- Conclusion, classifications, or results should ALWAYS appear last.
- Examples: Include high-quality examples if helpful, using placeholders [in brackets] for complex elements.
- What kinds of examples may need to be included, how many, and whether they are complex enough to benefit from placeholders.
- Clarity and Conciseness: Use clear, specific language. Avoid unnecessary instructions or bland statements.
- Formatting: Use markdown features for readability. DO NOT USE ``` CODE BLOCKS UNLESS SPECIFICALLY REQUESTED.
- Preserve User Content: If the input task or prompt includes extensive guidelines or examples, preserve them entirely, or as closely as possible. If they are vague, consider breaking down into sub-steps. Keep any details, guidelines, examples, variables, or placeholders provided by the user.
- Constants: DO include constants in the prompt, as they are not susceptible to prompt injection. Such as guides, rubrics, and examples.
- Output Format: Explicitly the most appropriate output format, in detail. This should include length and syntax (e.g. short sentence, paragraph, JSON, etc.)
- For tasks outputting well-defined or structured data (classification, JSON, etc.) bias toward outputting a JSON.
- JSON should never be wrapped in code blocks (```) unless explicitly requested.

The final prompt you output should adhere to the following structure below. Do not include any additional commentary, only output the completed system prompt. SPECIFICALLY, do not include any additional messages at the start or end of the prompt. (e.g. no "---")

[Concise instruction describing the task - this should be the first line in the prompt, no section header]

[Additional details as needed.]

[Optional sections with headings or bullet points for detailed steps.]

# Steps [optional]

[optional: a detailed breakdown of the steps necessary to accomplish the task]

# Output Format

[Specifically call out how the output should be formatted, be it response length, structure e.g. JSON, markdown, etc]

# Examples [optional]

[Optional: 1-3 well-defined examples with placeholders if necessary. Clearly mark where examples start and end, and what the input and output are. User placeholders as necessary.]
[If the examples are shorter than what a realistic example is expected to be, make a reference with () explaining how real examples should be longer / shorter / different. AND USE PLACEHOLDERS! ]

# Notes [optional]

[optional: edge cases, details, and an area to call or repeat out specific important considerations]
""".strip()

def generate_prompt(task_or_prompt: str):
completion = client.chat.completions.create(
model="gpt-4o",
messages=[
{
"role": "system",
"content": META_PROMPT,
},
{
"role": "user",
"content": "Task, Goal, or Current Prompt:\n" + task_or_prompt,
},
],
)

return completion.choices[0].message.content

More Than Just Text

And here’s the kicker—OpenAI didn’t stop at text. They’ve also rolled out a version of the Meta-Prompt for audio generation, with a few tweaks to better suit that medium. It’s all laid out in their documentation if you’re curious.

Looking ahead, OpenAI has hinted at even more advanced tools, like DSPy and Gradient Descent, potentially being added to the mix. So, if you’re already impressed, stay tuned—things are just getting started.