Advanced 30-40 minutes
Route Questions Between Your Specialist AI Twins
Design a practical routing layer for multiple specialist assistants so requests reach the right knowledge domain without unnecessary data sharing.
Recommended tools: ChatGPT, Claude, Google Gemini
Live Preview
ROLE
You are a multi-assistant systems architect.
OBJECTIVE
Design a routing specification for several specialist AI versions of my expertise.
CONTEXT
My specialist assistants are [ASSISTANT LIST]. Their knowledge domains are [DOMAIN DESCRIPTIONS]. Users submit [REQUEST TYPES]. Technical limitations are [PLATFORM LIMITATIONS].
INSTRUCTIONS
Create a routing taxonomy based on the work requested, not merely keywords. Define how to handle requests spanning multiple domains, requests outside every domain, and requests missing essential context. Specify the minimum information passed to each specialist and the information that must remain isolated. Design a handoff envelope containing the task, relevant evidence, constraints, and unresolved questions. Include a coordinator response format that keeps conflicting specialist recommendations visible. Provide six fictional routing examples, including one that should remain with the user. If automated routing is unavailable, supply an equivalent manual procedure.
CONSTRAINTS
Do not assume custom GPTs can call one another natively. Do not invent integrations, statistics, or successful executions. Flag assumptions and identify which steps require implementation outside the chat interface.
OUTPUT FORMAT
Routing matrix, handoff envelope, examples, manual fallback, and implementation checklist.
Open in:We copy the prompt first, then open the tool in a new tab. Results vary by tool.