Usage with LLMs & AI Agents¶

This page provides context-optimized documentation for Large Language Models (LLMs) and AI agents interacting with the edgework library.

System Prompt¶

When using an LLM to generate edgework code, provide the following context:

You are an expert in the edgework Python library for the NHL API. The library uses a main Edgework client class. - Seasons are often passed as strings "YYYY-YYYY" (e.g., "2023-2024") to the high-level API, but internally converted to integers (20232024). - Sorting uses parameters sort (snake_case property) and direction ("ASC" or "DESC"). - Models (Player, Team) are dynamic; access attributes via properties or get_attribute(). - Stats are fetched via client.skater_stats() or client.goalie_stats(), which return model instances containing lists of data entities.

API Map¶

Main Client: `Edgework`¶

from edgework import Edgework
client = Edgework()

Method	Signature	Description
`players`	`(active_only: bool = True) -> list[Player]`	Get list of players.
`get_teams`	`() -> list[Team]`	Get list of all teams.
`skater_stats`	`(season="YYYY-YYYY", report="summary", sort="points", limit=10, ...) -> SkaterStats`	Get skater statistics.
`goalie_stats`	`(season="YYYY-YYYY", report="summary", sort="wins", limit=10, ...) -> GoalieStats`	Get goalie statistics.
`team_stats`	`(season="YYYY-YYYY", report="summary", sort="points", ...) -> TeamStats`	Get team statistics.
`get_schedule`	`() -> Schedule`	Get current schedule.
`get_schedule_for_date`	`(date: str) -> Schedule`	Get schedule for "YYYY-MM-DD".
`get_roster`	`(team_code: str, season="YYYY-YYYY") -> Roster`	Get team roster.

Data Models¶

`Player`¶

Properties: name, full_name, position, sweater_number, current_team_abbr, age, height_formatted, weight, birth_city, birth_country.
Methods: get_attribute(name), fetch_data() (loads deep stats).

`Team`¶

Properties: name (e.g., "Toronto Maple Leafs"), abbrev ("TOR"), location ("Toronto"), full_name.
Methods: get_roster(), get_schedule(), get_stats().

`Roster`¶

Properties: players (list of Player), forwards, defensemen, goalies.
Methods: get_player_by_name(name), get_player_by_number(num).

One-Shot Examples¶

1. Find Top Scorers¶

from edgework import Edgework

client = Edgework()
stats = client.skater_stats(
    season="2023-2024",
    report="summary", 
    sort="points", 
    limit=5
)

for player in stats.players:
    print(f"{player.player_name} ({player.team_abbrevs}): {player.points} pts")

2. Analyze Team Roster¶

from edgework import Edgework

client = Edgework()
team_code = "EDM"
roster = client.get_roster(team_code)

print(f"{team_code} Roster:")
print(f"Forwards: {len(roster.forwards)}")
print(f"Defensemen: {len(roster.defensemen)}")
print(f"Goalies: {len(roster.goalies)}")

# Find specific player
mcdavid = roster.get_player_by_name("Connor McDavid")
if mcdavid:
    print(f"Captain: {mcdavid.name} (#{mcdavid.sweater_number})")

3. Get Schedule for Date¶

from edgework import Edgework

client = Edgework()
schedule = client.get_schedule_for_date("2023-11-18")

print(f"Games on {schedule.date}:")
for game in schedule.games:
    print(f"{game.away_team.abbrev} @ {game.home_team.abbrev} - {game.start_time}")

4. Advanced Stats Filtering¶

from edgework import Edgework

client = Edgework()
# Get top power play goal scorers
pp_leaders = client.skater_stats(
    season="2023-2024",
    report="powerPlay",
    sort="ppGoals",
    limit=10
)

for p in pp_leaders.players:
    print(f"{p.player_name}: {p.pp_goals} PPG")