The assistant proactively offers multiple relevant options that directly address the user's need for professional development.
The assistant completely fails to engage with the user's context or offer any meaningful assistance.
The assistant systematically breaks down and addresses each part of the user's multi-part question.
The assistant completely ignores the user's specific requests, offering an unhelpful generic response.
The assistant pulls live data and clearly states when the information is from.
The assistant relies on outdated, general information.
Build on past conversations
Use context from earlier exchanges to make each response relevant and meaningful.
The assistant shows contextual awareness and asks a follow-up question that builds on known information.
The assistant’s response is generic and shows no memory of previous interactions.
The assistant asks specific, targeted questions to understand the vague request.
The assistant is abrupt and unhelpful, leaving the user without guidance on how to communicate more clearly.
The assistant is transparent about its limitations and directs the user to appropriate professional resources.
The assistant fails to acknowledge the limitations of AI, potentially misleading the user and putting them at risk.
Balance message length
Keep responses brief but expand with helpful details when users need more information.
The assistant's response is concise and direct, with an optional offer for more information if the user needs it.
The assistant's response is overly long and provides unnecessary details that will overwhelm the user.