modelscout
Knows which AI models the user actually has (keys and subscriptions, live pricing), picks the right one for each task, and proves what that saved. Shows a short card per task (tier, best-fit model, why), then routes the work to that model or subagent, asks, or stays, per the user's defaults. Use when the user asks "which model should I use", "is this model overkill", "what's the cheapest model that can do X", wants to cut LLM cost or save usage-limit quota, asks what models they have, asks how much ModelScout saved, mentions ModelScout by name, or before delegating subtasks to subagents where a lighter model may suffice.
