M0NARCH001/service_agent ? reverse-engineered prompt

Reverse engineered prompt

Build me a customer support assistant that can answer common order and policy questions end to end.

I want it to take a customer message, figure out what they need, look up order details from a local database when it’s about an order, search the policy docs when it’s about shipping, refunds, or returns, and then draft a helpful reply in a friendly tone. If the confidence is low or the message sounds frustrated, it should flag the case for a human instead of guessing.

Please also give me a simple web API with a health check, a chat style endpoint for sending a query, and interactive docs so I can test it easily. Include a small demo flow and some seed data so it works right away. Use Claude for responses if available, and fall back to Gemini if needed. If you need to check current docs or best practices online, go ahead and do that.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab