Recommend Model Command

Use the model-recommender agent to help the user choose the best LLM for their task.

Process

1. Gather context:

   • If arguments provided: {{$ARGUMENTS}}
   • Pass any task description or requirements to the agent

2. Invoke the agent: Use the Task tool with subagent_type: "model-recommender" and pass:

   • The task description (if provided)
   • Any constraints mentioned (speed, cost, quality)
   • Special requirements (vision, JSON mode, etc.)

3. Present results: The agent will ask clarifying questions via AskUserQuestion, then return:

   • Top 3 model recommendations with match scores
   • Trade-off comparisons
   • Decision guide for choosing between options