To organize your prompts and make them modular and maintainable, you can store the text for Monday, start_of_month, and other conditions in separate dedicated files. Your prompts.json will then act as a mapping or configuration file, pointing to the corresponding files for each condition. Here’s how you can set it up:

1. File Organization

You can organize your files as follows:

prompts/ monday.txt start_of_month.txt other_day.txt prompts.json script.py

• monday.txt: Contains the narrative prompt for Mondays.
• start_of_month.txt: Contains the narrative prompt for the start of the month.
• other_day.txt: Contains the narrative prompt for other days.
• prompts.json: Maps the conditions to the corresponding text files.

2. Content of Each File

monday.txt

plaintext Analyze the data and focus on weekly trends, goals, and areas of improvement.

start_of_month.txt

plaintext Summarize the past month's performance and provide goals for the new month.

other_day.txt

plaintext Analyze today's data and focus on daily highlights and insights.

prompts.json

json { "monday": "prompts/monday.txt", "start_of_month": "prompts/start_of_month.txt", "other_day": "prompts/other_day.txt" }

3. Python Code to Load Prompts

Here’s how to dynamically load the text from the appropriate files:

```python import json

def load_prompt(condition, date, data): # Load the mapping from prompts.json with open("prompts.json", "r") as file: prompt_mapping = json.load(file)

# Determine the file path for the condition
prompt_file_path = prompt_mapping.get(condition)

# Read the content of the prompt file
with open(prompt_file_path, "r") as prompt_file:
    prompt_template = prompt_file.read()

# Substitute variables (if needed)
prompt = prompt_template.format(date=date, data=data)
return prompt

Determine the condition

from datetime import datetime

today = datetime.now() condition = ( "start_of_month" if today.day == 1 else "monday" if today.strftime("%A").lower() == "monday" else "other_day" )

Example variables

date = today.strftime("%Y-%m-%d") data = "Sales: $5M, Visitors: 10K"

Load the appropriate prompt

final_prompt = load_prompt(condition, date=date, data=data)

Print the final prompt

print(final_prompt) ```

4. Explanation of the Code

1. prompts.json as a Mapping:

2. Maps conditions like monday and start_of_month to their respective text file paths.

3. Dynamic File Loading:

4. Reads the appropriate text file based on the condition (e.g., Monday or start of the month).

5. Substituting Variables:

6. Uses .format(date=date, data=data) to inject dynamic variables into the prompt.

7. Condition Logic:

8. Determines whether it’s Monday, the start of the month, or another day using Python's datetime.

5. Example Output

For a Monday, assuming today's date is 2024-11-25, and data contains event information, the loaded prompt might look like this:

```plaintext Analyze the data and focus on weekly trends, goals, and areas of improvement.

Date: 2024-11-25 Data: Sales: $5M, Visitors: 10K ```

6. Advantages of This Approach

1. Separation of Concerns: Prompts are stored in individual files, keeping the main script clean and manageable.
2. Ease of Maintenance: Updating the prompt for a specific condition only requires editing the corresponding text file.
3. Scalability: Adding new conditions is as simple as creating a new text file and updating prompts.json.
4. Reusability: You can reuse the same structure for other projects requiring similar modular prompt management.

This approach provides a clean and scalable solution for managing prompts dynamically.