There are three levels in authoring propts:

First. Direct, simple instructions

For Example

data = {
    'prompt': """You are a cartoon artist. Please substitute the girl and cat in the cartoon image with the girl and cat from their respective images, following these requirements:
                1. Artfully substitute, freely adjust the size and position of the girl and cat.
                2. Keep the tone and feeling of the cartoon image.""",
    'n': 1,
    'size': '1024x1024'
}

Another Example:

logging.info("Sending request to OpenAI API.")
response = requests.post(api_url, headers=headers, files=files, data=data)
   
    response = openai.ChatCompletion.create(
    model="gpt-4",
    messages=[
        {"role": "system", "content": "Translate the following text from English to Chinese."},
        {"role": "user", "content": text_to_translate}
    ],
    max_tokens=8000,
    temperature=0.3,
    top_p=1.0,
    n=1,
    stop=None
)

Second. Applying Prompt Engineering Principles

content are from this page: https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/overview

Prompt generator

Be clear and direct

Use examples (multishot)

Let Claude think (chain of thought)

Use XML tags

Give Claude a role (system prompts

Prefill Claude’s response

Chain complex prompts

Long context tips

Third. The content of the prompt could change in each situation programtically

example:

https://github.com/anthropics/anthropic-cookbook/blob/main/skills/classification/guide.ipynb

prompt = textwrap.dedent("""
You will classify a customer support ticket into one of the following categories:
<categories>
    
</categories>

Here is the customer support ticket:
<ticket>
    
</ticket>

Use the following examples to help you classify the query:
<examples>
    
</examples>

First you will think step-by-step about the problem in scratchpad tags.
You should consider all the information provided and create a concrete argument for your classification.

Respond using this format:
<response>
    <scratchpad>Your thoughts and analysis go here</scratchpad>
    <category>The category label you chose goes here</category>
</response>
""").replace("", categories).replace("", X).replace("", rag_string)
response = client.messages.create( 
    messages=[{"role":"user", "content": prompt}, {"role":"assistant", "content": "<response><scratchpad>"}],
    stop_sequences=["</category>"], 
    max_tokens=4096, 
    temperature=0.0,
    model="claude-3-haiku-20240307"

Several SQL function that saved me a lot of time

Create A Simple Webapp