matchup-scout

name	matchup-scout
description	Scouts the current matchup when the user asks about their matchup, how they're doing this week, who they're playing, or their opponent's team. Pulls standings, matchup category breakdown, and opponent roster to give full context. Triggers on: 'matchup', 'how am I doing', 'who am I playing', 'opponent', 'this week', H2H category performance questions.

Matchup Scout

When the user asks about their current matchup, opponent, or weekly performance, follow these steps:

Step 1: Get League Context

Call get_standings to see where the user's team ranks and their record.

Step 2: Get Live Matchup

Call get_matchup with week=0 (current week) to get the category-by-category breakdown. This shows which categories you're winning, losing, or tied in.

Step 3: Scout the Opponent

From the matchup result, identify the opponent team name. Then:

• Call get_team_roster with the opponent's team name to see their full roster
• Call get_my_roster to see your own roster for comparison

Step 4: Check Memory (if available)

If the memory tools are available:

• Call get_matchup_history filtered by the opponent's name to see past results against them
• Call get_category_trends to understand which categories you've been strong or weak in recently

Step 5: Analyze and Recommend

Present a structured analysis:

1. Category Scoreboard: List all 14 categories with current status (Winning/Losing/Tied) and the values for both teams
2. Key Opponent Players: Highlight their best performers driving the categories they're winning
3. Strategic Recommendations:
   • Which categories to press (you're close to winning)
   • Which categories to protect (you're barely ahead)
   • Which categories to concede (too far behind, not worth chasing)
4. Roster Moves: If there are free agent pickups that could flip a close category, suggest them

Remember: 4 categories are reverse (lower wins): B_SO, ERA, WHIP, L. When these show a lower value for the user, that's GOOD.

Step 6: Save to Memory (if available)

If you observe notable category trends (e.g., consistently weak in SV, dominant in HR), call save_category_trend for each notable category.