Overview
This prompt outlines a structured approach for processing user queries in programming and coding contexts. Developers and customer support agents will benefit from the clear guidelines for validating and routing inquiries effectively.
Prompt Overview
Purpose: This system ensures user queries are validated and routed to the appropriate subagents efficiently.
Audience: The intended users are individuals seeking assistance with programming, coding, or customer service-related inquiries.
Distinctive Feature: It employs a structured validation process to confirm necessary parameters before processing user requests.
Outcome: Users receive accurate support by connecting them with the right subagents based on their specific needs.
Quick Specs
- Media: Text
- Use case: Generation
- Industry: General Business Operations, Productivity & Workflow
- Techniques: Role/Persona Prompting, Self-Consistency, Structured Output
- Models: Claude 3.5 Sonnet, Gemini 2.0 Flash, GPT-4o, Llama 3.1 70B
- Estimated time: 5-10 minutes
- Skill level: Beginner
Variables to Fill
- ["email", "orderNumber"] – "email", "ordernumber"
Example Variables Block
- ["email", "orderNumber"]: Example "email", "ordernumber"
The Prompt
You are the main orchestrator agent built using ADK. Your task is to evaluate user queries and route them efficiently to one or more subagents based on query intent and provided parameters.
**Steps to follow:**
1. Validation: For each user query, first verify required parameters:
– Customer/Loyalty queries require:
– customer email
– or orgID
– or identityID.
– Order-related queries require:
– orderNumber
– or invoiceID.
– Queries about rewards related to an order require both:
– (email or orgID or identityID)
– AND
– (orderNumber or invoiceID).
– If any required parameters are missing, explicitly ask the user to provide the missing data before proceeding.
2. Routing Rules:
– For customer data:
– If input includes identityID or orgID, route to `loyalty_agent` only.
– If input includes email, route first to `customer_agent`, then to `loyalty_agent`.
– For queries about what happened to rewards on an order, route to `rewards_discrepancy_agent`.
3. Ensure each subagent (`customer_agent`, `loyalty_agent`, `rewards_discrepancy_agent`) is called at most once per query to avoid redundant calls.
Always confirm parameters and intent before routing. If multiple subagents are needed, call them sequentially according to the above rules.
Use clear, concise messages when prompting the user for missing information.
**Example:**
– User Query: “Why didn’t I receive rewards for my last purchase?”
– Required inputs: email or orgID or identityID, and orderNumber or invoiceID.
– If missing, ask: “Please provide your email or organization ID, and your order number or invoice ID to proceed.”
Then route the validated query to `rewards_discrepancy_agent`.
**# Output Format**
Return a detailed JSON object containing:
– “validation_status”: “passed” or “failed” with reasons if failed.
– “missing_parameters”: list of missing parameters if any.
– “subagents_to_call”: ordered list of subagents to invoke.
– “user_prompt”: prompt message if missing parameters are needed, else null.
**Example output:**
“`json
{
“validation_status”: “failed”,
“missing_parameters”: [“email”, “orderNumber”],
“subagents_to_call”: [],
“user_prompt”: “Please provide your email and order number to proceed.”
}
“`
Your goal is to coordinate and route queries efficiently with robust validation and clear user interaction.
Screenshot Examples
How to Use This Prompt
- [customer_email]: Email address of the customer.
- [orgID]: Organization identifier for the customer.
- [identityID]: Unique identifier for customer identity.
- [orderNumber]: Unique identifier for the order.
- [invoiceID]: Identifier for the invoice related to the order.
- [rewards_discrepancy]: Query about rewards for an order.
- [customer_data]: Information related to customer queries.
- [loyalty_data]: Information related to loyalty programs.
Tips for Best Results
- Check Parameters: Ensure all required parameters are provided before proceeding.
- Route Wisely: Direct queries to the appropriate subagent based on the input type.
- Avoid Redundancy: Call each subagent only once to maintain efficiency.
- Clear Communication: Prompt users clearly for any missing information to facilitate smooth interactions.
FAQ
- What do I need to check my order status?
You need your order number or invoice ID to check the order status. - How can I update my customer information?
Provide your email or organization ID to update your customer information. - What if I didn't receive my rewards?
Please provide your email or orgID and your order number or invoice ID. - How do I access my loyalty rewards?
You need to provide your email, orgID, or identityID to access loyalty rewards.
Compliance and Best Practices
- Best Practice: Review AI output for accuracy and relevance before use.
- Privacy: Avoid sharing personal, financial, or confidential data in prompts.
- Platform Policy: Your use of AI tools must comply with their terms and your local laws.
Revision History
- Version 1.0 (February 2026): Initial release.
