fpl-mcp-server 0.1.7__tar.gz → 0.2.0__tar.gz

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/.github/workflows/lint.yml

@@ -21,7 +21,7 @@ jobs:
       - uses: actions/checkout@v6
 
       - name: Set up Python
-        uses: actions/setup-python@v5
+        uses: actions/setup-python@v6
         with:
           python-version: "3.13"
 

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fpl-mcp-server
-Version: 0.1.7
+Version: 0.2.0
 Summary: Fantasy Premier League MCP Server
 Project-URL: Homepage, https://github.com/nguyenanhducs/fpl-mcp
 Project-URL: Repository, https://github.com/nguyenanhducs/fpl-mcp
@@ -43,7 +43,7 @@ This MCP server provides comprehensive FPL analysis capabilities through:
 
 - **17 Interactive Tools** - Search players, analyze fixtures, compare managers, track transfers, and more
 - **4 Data Resources** - access to players, teams, and gameweeks bootstrap data
-- **8 Strategy Prompts** - Structured templates for squad analysis, transfer planning, chip strategy, and captain selection
+- **9 Strategy Prompts** - Structured templates for gameweek analysis, squad analysis, transfer planning, chip strategy, and captain selection
 - **Smart Caching** - 4-hour cache for bootstrap data to minimize API calls while keeping data fresh
 - **Fuzzy Matching** - Find players even with spelling variations or nicknames
 - **Live Transfer Trends** - Track the most transferred in/out players for current gameweek

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/README.md

@@ -12,7 +12,7 @@ This MCP server provides comprehensive FPL analysis capabilities through:
 
 - **17 Interactive Tools** - Search players, analyze fixtures, compare managers, track transfers, and more
 - **4 Data Resources** - access to players, teams, and gameweeks bootstrap data
-- **8 Strategy Prompts** - Structured templates for squad analysis, transfer planning, chip strategy, and captain selection
+- **9 Strategy Prompts** - Structured templates for gameweek analysis, squad analysis, transfer planning, chip strategy, and captain selection
 - **Smart Caching** - 4-hour cache for bootstrap data to minimize API calls while keeping data fresh
 - **Fuzzy Matching** - Find players even with spelling variations or nicknames
 - **Live Transfer Trends** - Track the most transferred in/out players for current gameweek

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/docs/tool-selection-guide.md

@@ -40,7 +40,7 @@ Interactive functions that perform specific FPL analysis tasks. All tools accept
 
 | Tool Name | Description | Key Parameters |
 |-----------|-------------|----------------|
-| `fpl_get_fixtures_for_gameweek` | All matches in a specific gameweek | `gameweek`, `format` |
+| `fpl_get_fixtures_for_gameweek` | All matches in a specific gameweek (basic or detailed) | `gameweek`, `detailed`, `format` |
 | `fpl_find_fixture_opportunities` | Find teams with easiest upcoming fixtures | `num_gameweeks`, `min_difficulty` |
 
 ### League & Manager Tools (4 tools)
@@ -94,6 +94,7 @@ Structured templates that guide analysis workflows. Prompts combine multiple too
 | `compare_players` | Side-by-side player comparison | `*player_names` | Choose between transfer targets |
 | `compare_managers` | Manager team comparison | `league_name`, `gameweek`, `*manager_names` | Analyze league rivals' strategies |
 | `find_league_differentials` | Find low-ownership differentials | `league_id`, `max_ownership` | Gain competitive advantage in mini-leagues |
+| `gameweek_analysis` | Detailed match reports & transfer implications | `gameweek` | Review completed gameweek stats and identify trends |
 
 ---
 
@@ -116,6 +117,7 @@ Structured templates that guide analysis workflows. Prompts combine multiple too
 
 - **Current gameweek?** → `fpl_get_current_gameweek` - Current or upcoming gameweek details
 - **Gameweek fixtures?** → `fpl_get_fixtures_for_gameweek` - All matches in a specific gameweek
+- **Gameweek review?** → `gameweek_analysis` - Comprehensive match analysis and transfer implications
 
 ### League & Manager Analysis
 

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/docs/usage-examples.md

@@ -17,6 +17,15 @@ Once configured, you can interact with the FPL MCP server through Claude Desktop
 "What teams have the best fixtures in the next 5 gameweeks?"
 "Show me Liverpool's upcoming fixtures"
 "Which fixtures are in gameweek 15?"
+"Show me detailed match reports for gameweek 12"
+
+## Gameweek Review
+
+```
+"Analyze the matches from last gameweek"
+"Who were the key performers in gameweek 5?"
+"Based on gameweek 10 matches, who should I buy?"
+```
 ```
 
 ## League & Manager Insights

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "fpl-mcp-server"
-version = "0.1.7"
+version = "0.2.0"
 description = "Fantasy Premier League MCP Server"
 readme = "README.md"
 requires-python = ">=3.13"

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/src/prompts/__init__.py

@@ -8,6 +8,7 @@ from ..tools import mcp
 from . import (
     captain_recommendation,  # noqa: F401
     chips,  # noqa: F401
+    gameweek_analysis,  # noqa: F401
     league_analysis,  # noqa: F401
     player_analysis,  # noqa: F401
     squad_analysis,  # noqa: F401

fpl_mcp_server-0.2.0/src/prompts/gameweek_analysis.py

@@ -0,0 +1,72 @@
+"""
+FPL MCP Prompts - Gameweek Analysis.
+
+Prompts to analyze detailed match reports and provide transfer advice for a gameweek.
+"""
+
+from ..state import store
+from ..tools import mcp
+from ..tools.fixtures import (
+    FindFixtureOpportunitiesInput,
+    GetFixturesForGameweekInput,
+    fpl_find_fixture_opportunities,
+    fpl_get_fixtures_for_gameweek,
+)
+
+
+@mcp.prompt()
+async def gameweek_analysis(gameweek: int | None = None) -> str:
+    """
+    Analyze detailed match reports for a specific gameweek.
+
+    Provides a comprehensive analysis of all matches in a gameweek, highlighting
+    key performers, tactical insights, and FPL implications based on detailed statistics.
+    Also suggests transfer targets based on upcoming fixtures and form.
+
+    Args:
+        gameweek: Gameweek number to analyze (defaults to current gameweek if None)
+    """
+    if gameweek is None:
+        gw_data = store.get_current_gameweek()
+        if not gw_data:
+            return "Error: Could not determine current gameweek."
+        gameweek = gw_data.id
+
+    # Fetch detailed fixtures
+    fixtures_input = GetFixturesForGameweekInput(gameweek=gameweek, detailed=True)
+    match_reports = await fpl_get_fixtures_for_gameweek(fixtures_input)
+
+    # Fetch simple fixture difficulty context for next 5 gameweeks
+    fixture_ops_input = FindFixtureOpportunitiesInput(num_gameweeks=5, max_teams=5)
+    fixture_opportunities = await fpl_find_fixture_opportunities(fixture_ops_input)
+
+    return f"""Analyze the following detailed match reports for Gameweek {gameweek}.
+
+**Match Reports:**
+{match_reports}
+
+**Upcoming Fixture Context (Next 5 GWs):**
+{fixture_opportunities}
+
+**Analysis Objectives:**
+1.  **Key Performers:** Identify the standout players based on goals, assists, and bonus points. Who is in top form?
+2.  **Tactical Insights:** Are there any noticeable trends? (e.g., high-scoring games, defensive masterclasses, specific teams dominating).
+3.  **FPL Implications:**
+    -   **Buy:** Comparison of assets who performed well.
+        -   **CRITICAL**: Cross-reference with Upcoming Fixture Context.
+        -   Generally prioritize players with good upcoming fixtures.
+        -   **EXCEPTION**: High-form elite players (fixture-proof) can be recommended even with tougher fixtures if their underlying stats are exceptional.
+        -   Flag 'Trap' assets: players in bad teams who scored once but have terrible fixtures/stats.
+    -   **Sell:** Notable failures or players who were benched/subbed early. Consider if it's a dip in form or a long-term issue.
+    -   **Watchlist:** interesting differentials or returning players.
+    -   **Analysis Depth**: When highlighting players, mention if their performance seems sustainable. Infer xG/xA quality from report descriptions if detailed stats suggest high involvement (e.g. many shots/key passes).
+4.  **Upcoming Outlook:** Based on this performance, who looks essential for the next gameweek?
+
+## 🔧 Tool Calls Strategy
+
+1.  **Deep Dive**: Use `fpl_get_player_details(player_name=...)` to check full history and upcoming fixtures for any player who catches your eye.
+2.  **Compare Options**: Use `fpl_compare_players(player_names=[...])` to decide between potential transfer targets.
+3.  **Check Manager**: If analyzing a specific rival, use `fpl_get_manager_by_team_id`.
+
+Provide a concise but deep analysis useful for an FPL manager making transfer decisions.
+"""

fpl_mcp_server-0.2.0/src/prompts/squad_analysis.py

@@ -0,0 +1,106 @@
+"""
+FPL MCP Prompts - Squad Performance Analysis.
+
+Prompts guide the LLM in analyzing squad performance over recent gameweeks
+using underlying metrics and regression analysis.
+"""
+
+from ..tools import mcp
+
+
+@mcp.prompt()
+def analyze_squad_performance(team_id: int, num_gameweeks: int = 5) -> str:
+    """
+    Analyze squad performance using xGI-based metrics and regression analysis.
+
+    This prompt guides the LLM to identify underperforming/overperforming players
+    using underlying stats (xG, xA, xGI) rather than retrospective points.
+
+    Args:
+        team_id: FPL team ID of the manager to analyze
+        num_gameweeks: Number of recent gameweeks to analyze (default: 5)
+    """
+    return f"""Analyze FPL squad performance for team ID {team_id} over the last {num_gameweeks} gameweeks.
+
+**OBJECTIVE: Create a PRO-LEVEL transfer strategy using Underlying Stats (xGI), Fixture Swings, and Chip Strategy.**
+
+---
+
+## 🏗️ **Step 1: Strategic Context (The "Manager's Eye")**
+
+Before analyzing players, assess the macro state of the squad:
+1.  **Financial Health:**
+    - Check `bank` balance. Can we afford luxury upgrades?
+    - Check `value` trends.
+2.  **Chip Status:**
+    - Which chips are available? (Wildcard, Free Hit, Bench Boost, Triple Captain)
+    - **Strategy:** If Wildcard is available and squad has >4 "issues", suggest Wildcard.
+3.  **Fixture Scan (Next 5 GW):**
+    - Identify teams with **Major Fixture Swings** (turning Good → Bad or Bad → Good).
+    - *Target:* Players from teams entering a "green run".
+    - *Avoid:* Players from teams entering a "red run".
+
+---
+
+## 📊 **Step 2: Player Performance Analysis (xGI Model)**
+
+Analyze each player using the **xGI Regression Model** (Output vs Expected):
+
+### **The Regression Framework:**
+- **xGI Delta** = `Actual G+A` - `Expected GI (xG + xA)`
+- **Interpretation:**
+    - **Huge Overperformance (+3.0+)**: *Elite Finisher* (e.g., Salah/Son) OR *Luck*? -> **HOLD** unless fixtures turn terrible.
+    - **Significant Underperformance (-2.0 to -3.0)**: *Unlucky*. If xGI is high (>0.5/90), **KEEP** or even **CAPTAIN**.
+    - **Poor Underlying (xGI < 0.2/90)**: *Ghosting*. Regardless of points, this is a **SELL** priority.
+
+### **Categorization:**
+- 🛡️ **Defensive Rocks**: Clean sheet potential + BP system appeal.
+- 🚜 **Workhorses**: Consistent xGI (0.4-0.6), nailed minutes.
+- 💣 **Explosive Differentials**: High xGI but low ownership (<10%).
+- 📉 **Dead Wood**: Low xGI + Bad Fixtures + Rotation Risk.
+
+---
+
+## 🔄 **Step 3: Transfer Planning (The "Next 3 Moves")**
+
+Instead of just "Buy X", provide a 3-Gameweek Plan:
+
+**Scenario A: The Surgery** (3+ Issues)
+- **GW{num_gameweeks + 1}:** Sell [Player A] -> Buy [Target A] (Reason: Fixture Swing)
+- **GW{num_gameweeks + 2}:** Roll Transfer / Sell [Player B]
+- **Long-term:** target [Premium Asset] in GW{num_gameweeks + 3}
+
+**Scenario B: The Luxury Move** (Squd is fine)
+- Upgrade specific position or build bank for future premium.
+
+**Prioritization Rules:**
+1.  **Injuries/Suspensions**: Immediate priority.
+2.  **Fixture Cliffs**: Selling players hitting a run of red fixtures.
+3.  **xGI Underperformers**: Selling players with low xGI (not just low points).
+
+---
+
+## 💡 **Recommendations & Targets**
+
+For each position (GKP, DEF, MID, FWD), recommend **Replacement Targets** based on:
+1.  **Fixture Swing**: Teams with Easiest Next 4 Games.
+2.  **Underlying Data**: Top xGI performers in those teams.
+3.  **Price Structure**: Must fit within budget (Bank + Sale Price).
+
+---
+
+## 🔧 **Tool Calls Strategy**
+
+1.  **Get Context**: `fpl_get_manager_by_team_id` (Squad, Bank, Rank) AND `fpl_get_manager_chips`.
+2.  **Scan Landscape**: `fpl_find_fixture_opportunities` (Identify teams to target).
+3.  **Analyze Squad**: `fpl_get_player_details` for current squad loop.
+4.  **Find Replacements**: `fpl_get_top_performers` (Cross-reference with Fixture Opportunities).
+5.  **Captaincy**: `fpl_get_captain_recommendations` for immediate GW.
+
+---
+
+## ⚠️ **Veteran Advice**
+- **Don't chase last week's points.** Look at who is *about* to score.
+- **Roll transfers** if the squad is healthy. 2 FTs is a superpower.
+- **Value < Points.** Don't hold a falling player just to save 0.1m if they aren't scoring.
+"""

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/src/tools/fixtures.py

@@ -22,6 +22,9 @@ class GetFixturesForGameweekInput(BaseModel):
     gameweek: int = Field(
         ..., description="Gameweek number to get fixtures for (1-38)", ge=1, le=38
     )
+    detailed: bool = Field(
+        default=False, description="Include detailed match stats (all available metrics)"
+    )
     response_format: ResponseFormat = Field(
         default=ResponseFormat.MARKDOWN,
         description="Output format: 'markdown' for human-readable or 'json' for machine-readable",
@@ -78,6 +81,7 @@ async def fpl_get_fixtures_for_gameweek(params: GetFixturesForGameweekInput) ->
     Args:
         params (GetFixturesForGameweekInput): Validated input parameters containing:
             - gameweek (int): Gameweek number between 1-38
+            - detailed (bool): Include detailed stats (default: False)
             - response_format (ResponseFormat): 'markdown' or 'json' (default: markdown)
 
     Returns:
@@ -111,26 +115,77 @@ async def fpl_get_fixtures_for_gameweek(params: GetFixturesForGameweekInput) ->
             result = {
                 "gameweek": params.gameweek,
                 "fixture_count": len(gw_fixtures_sorted),
-                "fixtures": [
-                    {
-                        "home_team": fixture.get("team_h_name"),
-                        "home_team_short": fixture.get("team_h_short"),
-                        "away_team": fixture.get("team_a_name"),
-                        "away_team_short": fixture.get("team_a_short"),
-                        "kickoff_time": fixture.get("kickoff_time"),
-                        "finished": fixture.get("finished"),
-                        "home_score": fixture.get("team_h_score")
-                        if fixture.get("finished")
-                        else None,
-                        "away_score": fixture.get("team_a_score")
-                        if fixture.get("finished")
-                        else None,
-                        "home_difficulty": fixture.get("team_h_difficulty"),
-                        "away_difficulty": fixture.get("team_a_difficulty"),
-                    }
-                    for fixture in gw_fixtures_sorted
-                ],
+                "fixtures": [],
             }
+
+            for fixture in gw_fixtures_sorted:
+                fixture_data = {
+                    "home_team": fixture.get("team_h_name"),
+                    "home_team_short": fixture.get("team_h_short"),
+                    "away_team": fixture.get("team_a_name"),
+                    "away_team_short": fixture.get("team_a_short"),
+                    "kickoff_time": fixture.get("kickoff_time"),
+                    "finished": fixture.get("finished"),
+                    "home_score": fixture.get("team_h_score") if fixture.get("finished") else None,
+                    "away_score": fixture.get("team_a_score") if fixture.get("finished") else None,
+                    "home_difficulty": fixture.get("team_h_difficulty"),
+                    "away_difficulty": fixture.get("team_a_difficulty"),
+                }
+
+                if params.detailed and fixture.get("stats"):
+                    stats = {}
+                    for stat in fixture.get("stats", []):
+                        if hasattr(stat, "model_dump"):
+                            stat_dict = stat.model_dump()
+                        elif hasattr(stat, "__dict__"):
+                            stat_dict = stat.__dict__
+                        else:
+                            stat_dict = stat
+
+                        identifier = stat_dict.get("identifier")
+                        if not identifier:
+                            continue
+
+                        stats[identifier] = {"h": [], "a": []}
+
+                        # Process home stats
+                        for item in stat_dict.get("h", []):
+                            if hasattr(item, "value"):
+                                value = item.value
+                                element = item.element
+                            else:
+                                value = item.get("value")
+                                element = item.get("element")
+
+                            stats[identifier]["h"].append(
+                                {
+                                    "name": store.get_player_name(element),
+                                    "value": value,
+                                    "element": element,
+                                }
+                            )
+
+                        # Process away stats
+                        for item in stat_dict.get("a", []):
+                            if hasattr(item, "value"):
+                                value = item.value
+                                element = item.element
+                            else:
+                                value = item.get("value")
+                                element = item.get("element")
+
+                            stats[identifier]["a"].append(
+                                {
+                                    "name": store.get_player_name(element),
+                                    "value": value,
+                                    "element": element,
+                                }
+                            )
+
+                    fixture_data["stats"] = stats
+
+                result["fixtures"].append(fixture_data)
+
             return format_json_response(result)
         else:
             output = [
@@ -159,6 +214,91 @@ async def fpl_get_fixtures_for_gameweek(params: GetFixturesForGameweekInput) ->
                     f"Difficulty: H:{fixture.get('team_h_difficulty')} A:{fixture.get('team_a_difficulty')}"
                 )
 
+                if params.detailed and fixture.get("stats"):
+                    stats_output = []
+                    # Filter for interesting stats or show all? Showing all as requested.
+                    # Common stats order: goals_scored, assists, own_goals, penalties_saved, penalties_missed, yellow_cards, red_cards, bonus, saves, bps
+                    priority_stats = [
+                        "goals_scored",
+                        "assists",
+                        "bonus",
+                        "yellow_cards",
+                        "red_cards",
+                        "saves",
+                    ]
+                    other_stats = []
+                    for s in fixture.get("stats", []):
+                        ident = s.identifier if hasattr(s, "identifier") else s.get("identifier")
+
+                        if ident and ident not in priority_stats:
+                            other_stats.append(ident)
+                    all_stats = priority_stats + other_stats
+
+                    for stat_name in all_stats:
+                        # Find stat data
+                        stat_data = None
+                        for s in fixture.get("stats", []):
+                            s_ident = (
+                                s.identifier if hasattr(s, "identifier") else s.get("identifier")
+                            )
+                            if s_ident == stat_name:
+                                stat_data = s
+                                break
+                        if not stat_data:
+                            continue
+
+                        # Convert to dict access if needed
+                        if hasattr(stat_data, "h"):
+                            h_items = stat_data.h
+                            a_items = stat_data.a
+                        else:
+                            h_items = stat_data.get("h", [])
+                            a_items = stat_data.get("a", [])
+
+                        if not h_items and not a_items:
+                            continue
+
+                        # Format stat line: Stat: Name (val), Name (val) (H) | Name (val) (A)
+                        h_strs = []
+                        for item in h_items:
+                            val = item.value if hasattr(item, "value") else item.get("value")
+                            elem = item.element if hasattr(item, "element") else item.get("element")
+                            h_strs.append(f"{store.get_player_name(elem)} ({val})")
+
+                        a_strs = []
+                        for item in a_items:
+                            val = item.value if hasattr(item, "value") else item.get("value")
+                            elem = item.element if hasattr(item, "element") else item.get("element")
+                            a_strs.append(f"{store.get_player_name(elem)} ({val})")
+
+                        stat_display = stat_name.replace("_", " ").title()
+                        if stat_name == "goals_scored":
+                            stat_display = "⚽ Goals"
+                        elif stat_name == "assists":
+                            stat_display = "🅰️ Assists"
+                        elif stat_name == "bonus":
+                            stat_display = "💎 Bonus"
+                        elif stat_name == "yellow_cards":
+                            stat_display = "🟨 Yellows"
+                        elif stat_name == "red_cards":
+                            stat_display = "🟥 Reds"
+                        elif stat_name == "saves":
+                            stat_display = "🧤 Saves"
+
+                        if h_strs or a_strs:
+                            line = f"   - **{stat_display}**: "
+                            if h_strs:
+                                line += f"{', '.join(h_strs)} ({home_name})"
+                            if h_strs and a_strs:
+                                line += " | "
+                            if a_strs:
+                                line += f"{', '.join(a_strs)} ({away_name})"
+                            stats_output.append(line)
+
+                    if stats_output:
+                        output.extend(stats_output)
+                        output.append("")  # Empty line separator
+
             result = "\n".join(output)
             truncated, _ = check_and_truncate(result, CHARACTER_LIMIT)
             return truncated

fpl_mcp_server-0.2.0/tests/test_gameweek_analysis_prompt.py

@@ -0,0 +1,57 @@
+from unittest.mock import AsyncMock, patch
+
+import pytest
+
+from src.prompts.gameweek_analysis import gameweek_analysis
+
+
+@pytest.mark.asyncio
+async def test_gameweek_analysis_prompt():
+    """Test that gameweek analysis prompt calls dependencies and formats output correctly."""
+
+    # Mock dependencies
+    with (
+        patch("src.prompts.gameweek_analysis.store") as mock_store,
+        patch(
+            "src.prompts.gameweek_analysis.fpl_get_fixtures_for_gameweek", new_callable=AsyncMock
+        ) as mock_get_fixtures,
+        patch(
+            "src.prompts.gameweek_analysis.fpl_find_fixture_opportunities", new_callable=AsyncMock
+        ) as mock_find_opportunities,
+    ):
+        # Setup mock returns
+        mock_store.get_current_gameweek.return_value.id = 5
+        mock_get_fixtures.return_value = "Detailed Match Reports Content"
+        mock_find_opportunities.return_value = "Top Teams: Arsenal, Liverpool"
+
+        # Call prompt
+        result = await gameweek_analysis(gameweek=None)
+
+        # Verify calls
+        mock_store.get_current_gameweek.assert_called_once()
+        mock_get_fixtures.assert_called_once()
+        mock_find_opportunities.assert_called_once()
+
+        # Check call arguments
+        fixtures_call_arg = mock_get_fixtures.call_args[0][0]
+        assert fixtures_call_arg.gameweek == 5
+        assert fixtures_call_arg.detailed is True
+
+        ops_call_arg = mock_find_opportunities.call_args[0][0]
+        assert ops_call_arg.num_gameweeks == 5
+        assert ops_call_arg.max_teams == 5
+
+        # Verify output content
+        assert "Gameweek 5" in result
+        assert "Detailed Match Reports Content" in result
+        assert "Upcoming Fixture Context" in result
+        assert "Top Teams: Arsenal, Liverpool" in result
+
+        assert "cross-reference" in result.lower()
+        assert "fixture-proof" in result
+        assert "trap" in result.lower()
+
+        # Check for tool strategy
+        assert "Tool Calls Strategy" in result
+        assert "fpl_get_player_details" in result
+        assert "fpl_compare_players" in result

fpl_mcp_server-0.2.0/tests/test_squad_analysis_prompt.py

@@ -0,0 +1,70 @@
+import os
+import re
+import sys
+
+
+def test_via_exec():
+    """Test the prompt content by executing the file content directly, bypassing imports."""
+    file_path = os.path.abspath(
+        os.path.join(os.path.dirname(__file__), "../src/prompts/squad_analysis.py")
+    )
+
+    with open(file_path) as f:
+        content = f.read()
+
+    # Remove the relative import that causes issues in isolation
+    content = re.sub(r"from \.\.tools import mcp", "", content)
+    # Remove the decorator
+    content = re.sub(r"@mcp\.prompt\(\)", "", content)
+
+    # Execute in isolated scope
+    local_env = {}
+    try:
+        exec(content, {}, local_env)
+    except Exception as e:
+        print(f"Failed to exec file: {e}")
+        sys.exit(1)
+
+    if "analyze_squad_performance" not in local_env:
+        print("Function analyze_squad_performance not found in executed code")
+        sys.exit(1)
+
+    analyze_func = local_env["analyze_squad_performance"]
+
+    # Test generation
+    team_id = 999999
+    prompt = analyze_func(team_id=team_id, num_gameweeks=5)
+
+    print("Prompt Preview:")
+    print("-" * 20)
+    print(prompt[:300] + "...")
+    print("-" * 20)
+
+    # Assertions
+    checks = [
+        f"team ID {team_id}",
+        "PRO-LEVEL transfer strategy",
+        "Strategic Context",
+        "Financial Health",
+        "Chip Status",
+        "fpl_get_manager_chips",
+        "fpl_find_fixture_opportunities",
+        "Scenario A: The Surgery",
+        "Scenario B: The Luxury Move",
+        "xGI Regression Model",
+    ]
+
+    failed = []
+    for check in checks:
+        if check not in prompt:
+            failed.append(check)
+
+    if failed:
+        print(f"❌ Verification Failed. Missing phrases: {failed}")
+        sys.exit(1)
+    else:
+        print("✅ Verification Passed! All Pro features present.")
+
+
+if __name__ == "__main__":
+    test_via_exec()

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/tests/test_tools_advanced.py

@@ -321,6 +321,36 @@ class TestAdvancedToolsIntegration:
             # Team 1 is Arsenal in mock_teams, short_name is ARS
             assert "ARS" in result
 
+    async def test_fpl_get_fixtures_for_gameweek_detailed(self, session_store, mock_fixtures_data):
+        """Test getting fixtures for gameweek with detailed stats."""
+        params = GetFixturesForGameweekInput(gameweek=2, detailed=True)
+
+        # Inject fixtures with stats
+        from src.models import FixtureData, FixtureStat, FixtureStatValue
+
+        fixtures = [FixtureData(**f) for f in mock_fixtures_data]
+        # Add stats to the first fixture
+        # Mocking Salah (ID 1) goal
+        fixtures[0].stats = [
+            FixtureStat(identifier="goals_scored", h=[FixtureStatValue(value=1, element=1)], a=[])
+        ]
+        session_store.fixtures_data = fixtures
+
+        with patch(
+            "src.tools.fixtures._create_client", new_callable=AsyncMock
+        ) as mock_create_client:
+            mock_client = AsyncMock()
+            mock_create_client.return_value = mock_client
+
+            result = await fpl_get_fixtures_for_gameweek(params)
+
+            assert "Gameweek 2" in result
+            assert "Goals" in result or "goals_scored" in result
+            # Should show player name (Salah from mock_players is ID 1)
+            # mock_players in conftest usually has ID 1 as Salah.
+            # session_store uses mock_players from conftest via fixture.
+            assert "Salah" in result
+
     async def test_fpl_analyze_team_fixtures_difficulty(self, session_store, mock_fixtures_data):
         """Test upcoming fixtures difficulty display using analyze_team_fixtures."""
         params = AnalyzeTeamFixturesInput(team_name="Arsenal", num_gameweeks=3)

{fpl_mcp_server-0.1.7 → fpl_mcp_server-0.2.0}/uv.lock

@@ -226,7 +226,7 @@ wheels = [
 
 [[package]]
 name = "fpl-mcp-server"
-version = "0.1.7"
+version = "0.2.0"
 source = { editable = "." }
 dependencies = [
     { name = "httpx" },

fpl_mcp_server-0.1.7/src/prompts/squad_analysis.py

@@ -1,136 +0,0 @@
-"""
-FPL MCP Prompts - Squad Performance Analysis.
-
-Prompts guide the LLM in analyzing squad performance over recent gameweeks
-using underlying metrics and regression analysis.
-"""
-
-from ..tools import mcp
-
-
-@mcp.prompt()
-def analyze_squad_performance(team_id: int, num_gameweeks: int = 5) -> str:
-    """
-    Analyze squad performance using xGI-based metrics and regression analysis.
-
-    This prompt guides the LLM to identify underperforming/overperforming players
-    using underlying stats (xG, xA, xGI) rather than retrospective points.
-
-    Args:
-        team_id: FPL team ID of the manager to analyze
-        num_gameweeks: Number of recent gameweeks to analyze (default: 5)
-    """
-    return f"""Analyze FPL squad performance for team ID {team_id} over the last {num_gameweeks} gameweeks.
-
-**OBJECTIVE: Identify transfer targets using xGI-based regression analysis, not retrospective points.**
-
----
-
-## 📊 **Performance Analysis Framework**
-
-For each player in the squad, analyze:
-
-### **1. Underlying Output Metrics (PRIMARY)**
-- **xGI/90 (Expected Goal Involvements per 90 min)**: Total xG + xA over last {num_gameweeks} GW, normalized per 90
-- **Minutes Played**: Total minutes + % of available minutes
-  *<60% = rotation risk*
-- **Games Played vs DNP**: Count starts, sub appearances, did not plays
-
-### **2. Regression Analysis (CRITICAL)**
-Calculate **xGI Delta** for each player:
-- `Actual G+A (last {num_gameweeks} GW)` MINUS `xG + xA (last {num_gameweeks} GW)`
-
-**Interpretation:**
-- **Positive Delta (+2 to +4)**: OVERPERFORMING → Likely to regress (sell candidate)
-- **Negative Delta (-2 to -4)**: UNDERPERFORMING → Due for improvement (keep/monitor)
-- **Near Zero (-1 to +1)**: Performing to expectation (stable)
-
-### **3. Player Categorization (xGI-Based)**
-
-Instead of arbitrary PPG thresholds, use xGI/90:
-
-- ⭐ **Elite Assets** (xGI/90 >0.6): Premium output, essential to keep
-  *Even if underperforming actual points (negative delta), underlying stats suggest improvement coming*
-
-- ✅ **Strong Contributors** (xGI/90 0.35-0.6): Reliable options, monitor for regression
-  *If positive delta >+2, consider selling before decline*
-
-- ⚠️ **Moderate Assets** (xGI/90 0.15-0.35): Acceptable for budget slots
-  *If negative delta <-2, potential buy-low candidates*
-
-- 🚨 **Underperformers** (xGI/90 <0.15): Transfer candidates
-  *Low underlying output + poor fixtures = priority sell*
-
-**Defenders/Goalkeepers:**
-- Use defensive contribution + clean sheet odds instead of xGI
-- xGC (Expected Goals Conceded) if available
-
----
-
-## 🔍 **Transfer Priority Analysis**
-
-### **For Each Underperformer (xGI/90 <0.15 OR injured):**
-
-1. **Regression Context:**
-   - If negative xGI delta: "Unlucky, but underlying stats still poor → Sell"
-   - If positive xGI delta: "Overperforming low xGI → Definitely sell before regression"
-
-2. **Availability Check:**
-   - Injured/Suspended → 🚨 **URGENT** (transfer immediately)
-   - DNP last 2 games → ⚠️ **HIGH** (rotation risk)
-   - Minutes <60% last {num_gameweeks} GW → 🟡 **MEDIUM**
-
-3. **Fixture Difficulty (Next 4 GW):**
-   - Avg FDR >3.5 → Poor fixtures exacerbate low xGI
-   - Avg FDR <2.5 → Fixtures can't help poor underlying stats
-
-4. **Ownership Context:**
-   - If template player (>30% ownership): May need to hold for rank protection
-   - If differential (<10% ownership): Easy sell, minimal rank impact
-
-5. **Transfer Recommendation:**
-   - Provide urgency level: 🚨 URGENT / ⚠️ HIGH / 🟡 MEDIUM / 🟢 LOW
-   - Suggest xGI-based replacement targets from `fpl_get_top_performers`
-
----
-
-## 📈 **Squad Health Summary**
-
-1. **Player Counts by Category:**
-   - Elite Assets (xGI/90 >0.6): [X] players
-   - Strong Contributors (xGI/90 0.35-0.6): [X] players
-   - Moderate Assets (xGI/90 0.15-0.35): [X] players
-   - Underperformers (xGI/90 <0.15): [X] players
-
-2. **Regression Risk Summary:**
-   - Players with xGI Delta >+2 (overperforming): [List names] → Potential sell targets
-   - Players with xGI Delta <-2 (underperforming): [List names] → Monitor for improvement
-
-3. **Priority Transfer Target:**
-   - **Player Name** (xGI/90: [X.XX], xGI Delta: [+/-X.X], Fixtures: [avg FDR])
-   - **Reason:** [injury / low xGI / rotation / regression risk]
-   - **Urgency:** [URGENT/HIGH/MEDIUM/LOW]
-
-4. **Overall Squad Health:**
-   - Healthy (8+ strong contributors) / Moderate (5-7) / Poor (<5)
-
----
-
-## 🔧 **Tool Calls**
-
-Use these tools and resources:
-1. `fpl_get_manager_by_team_id(team_id={team_id})` → Current squad composition
-2. `fpl_get_top_performers(num_gameweeks={num_gameweeks})` → Benchmark against top xGI players
-3. For each player:
-   - `fpl://player/{{{{player_name}}}}/summary` → xG, xA, xGI, minutes, fixtures
-4. `fpl://bootstrap/players` → Ownership %, price, transfer trends
-
----
-
-## ⚠️ **Critical Rules**
-
-1. **NEVER categorize by PPG alone** → Use xGI/90 for attackers, defensive contribution for defenders
-2. **ALWAYS calculate xGI Delta** → Regression context is critical for sell decisions
-3. **Account for ownership** → Template players need more justification to sell
-4. **Prioritize injured/unavailable** → These are auto-sell regardless of xGI
-"""