LLM Response Record

Generated on:
ID:
Created at:

---

🤖 Query

---

💬 Response

---

🔧 Model Information

Provider:
Model:

---

📄 Raw Data

Response JSON

{{RESPONSE_JSON}}

---

Template Usage Instructions

To use this template, replace the placeholders with your actual data:

Placeholders to Replace:

• → Date in YYYY-MM-DD format (e.g., 2025-10-11)
• → Human-readable generation timestamp (e.g., 2025-10-11 12:19:56)
• → The record ID (e.g., 68dd1f52292b8087b556615a)
• → ISO timestamp (e.g., 2025-10-01T12:32:18.118141300Z)
• → The query or prompt sent to the LLM
• → The LLM's response text
• → The LLM provider (e.g., Perplexity, OpenAI)
• → The specific model used (e.g., sonar, gpt-4)
• → The raw JSON response from the API (formatted properly)

Python Helper Function

If you want to automate this process, here's a Python function you can add to your script.py:

import json
from datetime import datetime

def format_llm_response_record(record_data):
    """
    Format LLM response record into styled markdown
    
    Args:
        record_data (dict): Dictionary containing:
            - id: Record ID
            - created_at: ISO timestamp
            - query: The query/prompt
            - response: The LLM response text
            - provider: LLM provider name
            - model: Model name
            - response_json: Raw JSON response (optional)
            - generated_date: Human readable date (optional)
    """
    
    # Format date
    created_dt = datetime.fromisoformat(record_data['created_at'].replace('Z', '+00:00'))
    date_str = created_dt.strftime('%Y-%m-%d')
    generated_str = record_data.get('generated_date', created_dt.strftime('%Y-%m-%d %H:%M:%S'))
    
    # Handle response_json
    response_json = record_data.get('response_json', {})
    if isinstance(response_json, str):
        response_json = json.loads(response_json)
    
    # Build template
    template = f"""---
title: LLM Response Record
description: LLM response record with query, response, and model information
date: {date_str}
---

# LLM Response Record

**Generated on:** {generated_str}  
**ID:** {record_data['id']}  
**Created at:** {record_data['created_at']}

---

## 🤖 Query

{record_data['query']}

---

## 💬 Response

{record_data['response']}

---

## 🔧 Model Information

**Provider:** {record_data['provider']}  
**Model:** {record_data['model']}

---

## 📄 Raw Data

### Response JSON
```json
{json.dumps(response_json, indent=2)}
```"""
    
    return template

This template and helper function will ensure consistent formatting for all your LLM response records, making them much more readable and professional-looking. You can easily copy the template and fill in the placeholders manually, or use the Python function to automate the process.