sportsdata-mcp

DFS fantasy-points projections + salaries/ownership for a site + slate. Returns: {event_name, last_updated, projections:[{dg_id, player_name, proj_points, salary, proj_ownership}]} Auth: needs your own key in DATAGOLF_KEY.

Current event field — players entered, tee times, withdrawals, current round. Returns: {event_name, event_id, course_name, current_round, date_start, date_end, field:[{dg_id, player_name, ...tee times}]} Auth: needs your own key in DATAGOLF_KEY.

List of historical events with DFS salary/ownership/points data for a site. Returns: [{event_id, event_name, calendar_year, date, dk_salaries, dk_ownerships}] (top-level array) Auth: needs your own key in DATAGOLF_KEY.

Historical DFS points + salary + ownership per player for a past event. Returns: {event_id, event_name, event_completed, season, dfs_points:[{dg_id, player_name, fin_text, finish_pts, total_pts, salary, ownership}]} Auth: needs your own key in DATAGOLF_KEY.

List of historical events that have raw round-level data available (event_id + year to pass to the rounds feed). Returns: [{event_id, event_name, tour, calendar_year, date, sg_categories}] (top-level array) Auth: needs your own key in DATAGOLF_KEY.

Historical matchup / 3-ball odds for a past event from one sportsbook. Returns: {book, event_id, event_name, event_completed, matchups:[{p1_player_name, p2_player_name, open_odds, close_odds, ...}]} Auth: needs your own key in DATAGOLF_KEY.

List of historical events that have archived bookmaker odds available. Returns: [{event_id, event_name, calendar_year, archived_preds, ...}] (top-level array) Auth: needs your own key in DATAGOLF_KEY.

Historical opening/closing outright odds for a past event from one sportsbook. Returns: {book, event_id, event_name, event_completed, outrights:[{dg_id, player_name, open_odds, close_odds, ...}]} Auth: needs your own key in DATAGOLF_KEY.

Historical event-level results for a past tournament — finish position, earnings, FedExCup points and DG Points per player. Returns: {event_id, event_name, year, season, event_completed, event_stats:[{dg_id, player_name, fin_text, earnings, fec_points, dg_points}]} Auth: needs your own key in DATAGOLF_KEY.

List of tournaments with historical event-level results data (finishes/earnings/points) — the id lookup for datagolf_hist_results. Returns: [{event_id, event_name, calendar_year, date}] (top-level array) Auth: needs your own key in DATAGOLF_KEY.

Historical round-by-round scoring + strokes-gained for every player in a past event. Returns: {event_id, event_name, event_completed, scores:[{dg_id, player_name, fin_text, rounds:[{round_num, score, sg_total, sg_putt, ...}]}]} Auth: needs your own key in DATAGOLF_KEY.

Live (in-play) model predictions during a tournament — current position/score + updated win probabilities. Returns: {data:[{dg_id, player_name, current_pos, current_score, R1, R2, R3, R4, make_cut, win, ...}]} Auth: needs your own key in DATAGOLF_KEY.

Live per-hole scoring distributions for the in-progress event (avg score, birdie/bogey rates by hole, AM/PM wave). Returns: {courses:[{course_code, course_key, rounds:[{round_num, holes:[{hole, par, avg_score, afternoon_wave:{...}, morning_wave:{...}}]}]}]} Auth: needs your own key in DATAGOLF_KEY.

Live strokes-gained breakdown for every player during PGA Tour events (raw, or relative to the model's pre-round expectations). Returns: {event_name, current_round, last_update, strokes_gained_values, data:[{dg_id, player_name, pos, score, thru, today, R1:{...sg}, R2, R3, R4}]} Auth: needs your own key in DATAGOLF_KEY.

Live per-player tournament stats (strokes-gained by category, accuracy, etc.) for the requested round. Returns: {event_name, course_name, last_updated, live_stats:[{dg_id, player_name, sg_total, sg_ott, sg_app, ...}]} Auth: needs your own key in DATAGOLF_KEY.

Tournament / round / 3-ball matchup odds across sportsbooks, plus Data Golf's model line. Returns: {event_name, market, last_updated, match_list:[{p1_player_name, p2_player_name, odds:{datagolf, bet365, pinnacle, ...}}]} Auth: needs your own key in DATAGOLF_KEY.

All possible tournament matchup pairings for the current event with Data Golf's model odds (every player-vs-player price). Returns: {event_name, last_update, pairings:[{p1_player_name, p2_player_name, p3_player_name, ...model odds}]} Auth: needs your own key in DATAGOLF_KEY.

Outright (win / top-N / make-cut) odds for the current event across ~13 sportsbooks, plus Data Golf's model line. Returns: {event_name, last_updated, books_offering:[...], odds:[{dg_id, player_name, datagolf:{...}, bet365, pinnacle, draftkings, fanduel, ...}]} Auth: needs your own key in DATAGOLF_KEY.

Per-player skill decomposition for the current event — how each skill (driving, approach, putting, course-fit) contributes to the prediction. Returns: {event_name, course_name, players:[{dg_id, player_name, ...skill/course-fit components}]} Auth: needs your own key in DATAGOLF_KEY.

Full player list with Data Golf ids (dg_id), name, country, amateur flag. Returns: [{dg_id, player_name, country, country_code, amateur}] (top-level array) Auth: needs your own key in DATAGOLF_KEY.

Pre-tournament model predictions — win / top-5/10/20 / make-cut probabilities per player. Returns: {baseline:[{dg_id, player_name, win, top_5, top_10, top_20, make_cut}], baseline_history_fit:[...]} Auth: needs your own key in DATAGOLF_KEY.

Archived pre-tournament predictions for a past event (what the model said beforehand). Returns: {baseline:[{dg_id, player_name, fin_text, win, make_cut, first_round_leader, top_10}]} Auth: needs your own key in DATAGOLF_KEY.

Data Golf player rankings (top ~500) with skill estimate and OWGR rank. Returns: {last_updated, rankings:[{datagolf_rank, owgr_rank, dg_id, player_name, dg_skill_estimate, country}]} Auth: needs your own key in DATAGOLF_KEY.

Tour schedule — the season's events with dates, courses and locations. Returns: {schedule:[{event_id, event_name, course, course_key, location, country, latitude, longitude, start_date}]} Auth: needs your own key in DATAGOLF_KEY.

Player skill ratings — strokes-gained components (off-tee, approach, around-green, putting) + driving acc/dist. Returns: {last_updated, players:[{dg_id, player_name, sg_total, sg_ott, sg_app, sg_arg, sg_putt, driving_acc, driving_dist}]} Auth: needs your own key in DATAGOLF_KEY.

Contentful CMS entries (promotions, major-event nav) via the www CDN proxy. Returns: {sys, total, skip, limit, items:[{sys, fields}], includes:{}} Example: Active promotions {"content_type": "promotions", "limit": 20} Auth: none needed.

Settlement rules for every named market, indexed by rule id. Returns: {market_rules:{<rule_id>:{id, name, description}}} Auth: none needed.

Join table between market types and market-type groups (dedup by pair). Returns: {market_type_group_map:[{market_type_id, market_type_group_id, priority}]} Auth: none needed.

Market-tab group definitions for a sport category (lower priority renders first). Returns: {market_type_group:[{id, category_id, name, priority, type}]} Auth: none needed.

Featured slider events (homepage hero carousel). Returns: {items:[{id, title, url, event_id, event:{competition_id, event_start}}]} Auth: none needed.

Call any of Entain's 127 persisted GraphQL operations against api.ladbrokes.com.au/gql/router by name + variables. Hashes are managed server-side; a PERSISTED_QUERY_NOT_FOUND (gateway APQ-cache eviction) is self-healed automatically by re-POSTing the stored query document. If it still surfaces, run `sportsdata-mcp refresh-hashes entain`. GraphQL IDs are type-prefixed (RacingRace:<uuid>, SportingEvent:<uuid>): variable type ID! wants the prefixed form, UUID! wants the bare uuid. Read entain://graphql/operations for the full op list + variable signatures. Returns: (JSON object) Auth: none needed.

SEO metadata (page title) for a given URL path. Returns: {metadata:{title, url}} (empty {} when no override) Auth: none needed.

Navigation quick-links (racing/sports nav tiles). Returns: {quick_links:[{id, root_category_id, category_id, url, title, icon, priority}]} Example: Racing quick-links {"filter": {"type": "Racing"}} Auth: none needed.

Legacy v1 racing RPC selector (future-markets races feed). Returns: {status, data:{races:{<race_id>:{name, advertised_start:{seconds}}}}} Auth: none needed.

All race meetings + races for one date (normalised UUID-keyed tables). Returns: {compounds, domestic_countries, meetings:{}, races:{}, venues:{}} Example: Melbourne card for a given day {"date": "<today>", "timezone": "Australia/Melbourne"} Auth: none needed.

Next races about to jump, grouped per racing category. Returns: {category_race_map:{}, race_summaries:{<race_id>:{race_name, meeting_name, advertised_start, race_form}}} Auth: none needed.

Full priced racecard for one race — entrants, fixed-odds fluctuations, form. Returns: {status, data:{races:{}, markets:{}, prices:{}, entrants:{}, price_fluctuations:{<entrant_id>:[floats — LAST is the live fixed win price]}, meetings:{}}} Example: One priced racecard {"id": "7f553143-1ed4-4ef8-a622-46c7563e6c83"} Auth: none needed.

Racing search facets (barrier/country/jockey/trainer buckets); optional full-text. Returns: {facets:{barrier:{buckets}, country:{buckets}, jockey:{buckets}, trainer:{buckets}}} Auth: none needed.

Complete event card — every market, selection and price for one sport event. Returns: {entrants:{}, events:{}, markets:{}, prices:{}, regions:{}, market_type_groups:{}} Example: One NBA event card {"id": "339e26d0-72a8-49bc-a85f-a2d02c0a1a70"} Auth: none needed.

Bulk events + markets + prices for one or more sport categories. Returns: {events:{}, markets:{}, prices:{}, entrants:{}, next_events:[], regions:{}} Auth: none needed.

Racing live-video channels (HLS .m3u8 URLs; verify token expires within minutes). Returns: {channels:[{id, name, url}]} Auth: none needed.

Competitions and tours. Returns: {status:'ok', response:{items:[{cid, title, abbr, category, game_format, status, season, datestart, dateend, total_matches, total_teams}], total_pages}} — SHAPE FROM VENDOR DOCS. NOTE: this shape is from the vendor's documentation and has NOT been verified against a live response (we hold no key for this provider). Treat it as approximate — inspect the actual payload before relying on a field name. Example: Live competitions {"status": "live"} Auth: needs your own key in ENTITYSPORT_TOKEN.

Ball-by-ball commentary for an innings — the reason to choose this provider over cricketdata. Returns: {status:'ok', response:{commentaries:[{event:'ball'|'overend'|'wicket', over, ball, batsman_id, bowler_id, score, run_str, commentary, noball, wide, byes, legbyes, six, four}]}} — SHAPE FROM VENDOR DOCS. Rows include NON-BALL events ('overend'), so filter on `event` before counting deliveries. NOTE: this shape is from the vendor's documentation and has NOT been verified against a live response (we hold no key for this provider). Treat it as approximate — inspect the actual payload before relying on a field name. Example: An innings' commentary {"matchId": 12345, "inningsId": 1} Auth: needs your own key in ENTITYSPORT_TOKEN.

One match with squads, toss and current state. Returns: {status:'ok', response:{match_id, title, format_str, status_str, teama, teamb, venue, toss, umpires, referee, players:[…], equation, live}} — SHAPE FROM VENDOR DOCS. NOTE: this shape is from the vendor's documentation and has NOT been verified against a live response (we hold no key for this provider). Treat it as approximate — inspect the actual payload before relying on a field name. Example: One match {"matchId": 12345} Auth: needs your own key in ENTITYSPORT_TOKEN.

Full innings-by-innings scorecard. Returns: {status:'ok', response:{innings:[{iid, number, name, batsmen:[{name, batsman_id, runs, balls_faced, fours, sixes, strike_rate, how_out}], bowlers:[{name, overs, maidens, runs, wickets, econ}], extras, equations:{runs, wickets, overs, runrate}, fows:[…], did_not_bat:[…]}]}} — SHAPE FROM VENDOR DOCS. `fows` is the fall-of-wickets list. NOTE: this shape is from the vendor's documentation and has NOT been verified against a live response (we hold no key for this provider). Treat it as approximate — inspect the actual payload before relying on a field name. Example: One match's scorecard {"matchId": 12345} Auth: needs your own key in ENTITYSPORT_TOKEN.

Cricket matches — live, upcoming and completed. Returns: {status:'ok', response:{items:[{match_id, title, short_title, subtitle, format_str:'T20'|'ODI'|'Test', status, status_str, game_state_str, teama:{team_id, name, short_name, scores_full, scores, overs}, teamb:{…}, date_start, venue, toss:{text, winner, decision}}], total_items, total_pages}} — SHAPE FROM VENDOR DOCS. `scores_full` is the display string ('187/4 (20)') and `scores` the bare runs — do not parse the first when you want the second. NOTE: this shape is from the vendor's documentation and has NOT been verified against a live response (we hold no key for this provider). Treat it as approximate — inspect the actual payload before relying on a field name. Example: Live matches {"status": 3} Auth: needs your own key in ENTITYSPORT_TOKEN.

Gateway to the ESPN CDN core live feed (cdn.espn.com). Fast, lightly-cached scoreboard/game/boxscore/playbyplay JSON. The path slug is the LEAGUE, not the sport (`nfl`, `nba`, `mlb`, `college-football`, `mens-college-basketball`) — for soccer use the competition slug directly (`eng.1`, `esp.1`, `uefa.champions`). The CDN covers ESPN's front-page leagues only (no `nhl`). gameId comes from the scoreboard op. Every request carries ?xhr=1 (added by default). Browse espn://cdn/operations. Returns: (JSON object) Auth: none needed.

Gateway to the ESPN core data model (sports.core.api.espn.com). The deepest, most canonical surface: athletes + statistics + career logs, per-event odds / win-probabilities / plays / situation / broadcasts / predictor / power-index, per-competitor line-scores + statistics, season teams/coaches/draft/futures, venues, leaders, rankings, franchises and coaches. NOTE the path uses `leagues/{league}` (plural). Supply `operation` + `path_params` (sport, league, plus eventId/competitionId/athleteId/year as needed). Most responses use ESPN's $ref-linked envelope ({count, items:[{$ref}]}); follow the refs for detail. Browse every operation in the espn://core/operations resource. Returns: (JSON object) Auth: none needed.

Full summary for one game: box score, scoring plays, drives/play-by-play, leaders, win-probability and odds. `event` is an id from espn_scoreboard. No example here on purpose — it needs a live event id, so the doctor probes espn_scoreboard instead. Returns: {boxscore:{teams, players}, plays:[...], scoringPlays:[...], leaders:[...], winprobability:[...]} Auth: none needed.

League news feed: recent articles (headline, description, links, images, related athletes/teams) for one league. Returns: {header, articles:[{headline, description, published, links, categories}]} Example: Latest NFL news. {"sport": "football", "league": "nfl"} Auth: none needed.