MCP App Store
by KeenEthics
FantasyPros: Fantasy Advice icon

FantasyPros: Fantasy Advice MCP Server for ChatGPT

By FantasyPros

Overview

Bring FantasyPros into ChatGPT for personalized fantasy football and fantasy baseball advice! Powered by the FantasyPros Expert Consensus Rankings and the advice tools behind My Playbook. Connect your FantasyPros account to analyze your synced leagues, compare players, evaluate trades, find waiver targets, check matchups, and get start/sit guidance tailored to your team. Rankings and research tools work without a synced league; league-specific advice uses the leagues connected to your FantasyPros account.

How to connect

  1. ChatGPT: Open the FantasyPros: Fantasy Advice listing in the ChatGPT app directory and choose Connect. ChatGPT listing FantasyPros: Fantasy Advice exposes 28 tools here.

Tools

get_adp

ChatGPT
Get average draft position (ADP) — where players are actually being drafted across real fantasy drafts. Each row carries the ADP value (average pick), overall rank, position rank, and bye week. Useful for draft prep and for spotting value gaps between expert rankings (get_ecr) and market cost (this tool).

get_current_starters

ChatGPT
Get current starters and bench players from matchup lineup data. Returns status='league_not_found' when the league key is not recognized upstream (deleted, expired, or mistyped) — re-resolve keys via get_leagues.

get_ecr

ChatGPT
Get consensus expert rankings (ECR) for NFL or MLB.

get_keeper_value

ChatGPT
Evaluate keeper decisions: weigh each player's draft value against their keeper cost. Keepers are ranked by value (an auction-value delta). Results are framed in the user's own units: auction leagues return values and deltas in their league's budget dollars; pick-cost leagues return the player's estimated true draft slot (their overall cheatsheet rank) plus the approximate draft picks gained by keeping them. Values come from the league's DraftWizard cheatsheet (which bakes in the league's scoring).

get_leaders

ChatGPT
Get leaderboard results for a stat category.

get_league_settings

ChatGPT
Get league scoring, dynasty, and roster settings, plus the team directory (teams: [{team_id, team_name, is_user_team}]) — use it to resolve the exact team_name/team_id for get_roster. In raw_settings.scoring_system, position-suffixed types (RecRB, RecWR, RecTE) are the effective per-position values and replace the generic type (Rec) for that position — they do NOT stack with it.

get_leagues

ChatGPT
Get the authenticated user's synced leagues. Returns a compact summary per league: the league `key` (the id to pass to league-scoped tools), league/team names, host, size, scoring, and league-type flags.

get_mlb_lineups

ChatGPT
Get MLB lineups and probable pitchers from the FantasyPros v2 API. rank_vbr on pitcher rows is a league-agnostic (site default scoring) value-based rank from this endpoint. It is NOT comparable to the rank_vbr from get_streaming_pitchers, which is computed by a different upstream service with the user's league context — for league decisions prefer the get_streaming_pitchers ordering.

get_mlb_player_ownership

ChatGPT
Find whether a player is owned in an MLB league.

get_nfl_depth_charts

ChatGPT
Get NFL depth chart data by team or player. In season, each player's ecrRank is this week's expert consensus rank (the rest-of-season rank for a team on bye), the same number the FantasyPros Depth Charts page shows. Out of season it is the draft rank.

get_player_comparisons

ChatGPT
Compare 2-4 players for weekly/ROS/draft/dynasty formats.

get_player_ownership

ChatGPT
Check whether an NFL player is rostered or available across the user's leagues. Use this whenever the user asks if a player is owned, taken, available, on the waiver wire, or a free agent. Scans every team in each of the user's synced NFL leagues and reports, per league, which team owns the player (flagging the user's own team) or that they're an available free agent. This is NOT limited to the user's own roster — prefer it over get_roster for any ownership/availability question.

get_player_stats

ChatGPT
Get NFL stats for one or more players. Covers individual defensive players as well as offense. For IDP, solo and assisted tackles ARE returned, as def_tackle and def_assist, alongside def_sack, def_int, def_ff and def_fr — check the response before telling a user a defensive stat is unavailable. Tackles for loss, passes defended and QB hits are the ones genuinely absent. For IDP rows the points* fields are a generic default, not the league's scoring; the response says so in instructions. Never quote them as the player's fantasy points.

get_projections

ChatGPT
Get expert-consensus player projections (NFL or MLB) from FantasyPros. NFL rows include standard, half-PPR, and full-PPR points plus the per-position stat lines; MLB rows carry hitter or pitcher stat lines.

get_roster

ChatGPT
Get one or all team rosters for an NFL or MLB league. With no selector, returns the authenticated user's own team. Pass team_id for an exact team-id match (it takes precedence over team_name), or pass team_name for an exact match (case-insensitive; no fuzzy matching). If you only have an approximate name, first resolve the exact one from get_league_settings (teams) — or from the available_teams list this tool returns on a miss. Pass all_teams=true to return every team; it cannot be combined with a team selector.

get_schedule_and_strength

ChatGPT
Get team schedules and strength-of-schedule data. Strength-of-schedule exists for offensive positions and for team defenses (DST) only. IDP leagues have no per-defender SOS: a DST number describes how opposing offenses are expected to fare against that team's entire defense, and says nothing about an individual linebacker or edge rusher. Never answer a question about one IDP player with a DST schedule. The response echoes the effective SOS window (sos_window / per-team "weeks" label), the current week, and season_state (current_week is 0 in the offseason). Every team carries its bye_week, so pass all the teams in a lineup to check byes for any week, including a later one.

get_streaming_pitchers

ChatGPT
Get the best unrostered streaming pitcher options for a day. Results are ordered by rank_vbr (lower is better), a value-based rank computed with this league's context. It is NOT comparable to the rank_vbr from get_mlb_lineups, which comes from a different, league- agnostic upstream source — the same pitcher gets different numbers.

injury_status

ChatGPT
Get injury reports and the last two weeks of FantasyPros news for players. Player lookup requires full first and last names; expand any shorthand from the conversation (e.g. 'Sadiq' -> 'Kenyon Sadiq') before calling. practice_details is the current injury report. player_news is FantasyPros news for these players from the last 14 days, newest first, each item with its published time (UTC) and a fantasypros.com url. Say when the news you cite was published. The injury report covers INJURIES ONLY -- it is not a roster or status check. A player with no practice information is simply absent from the injury feed, which is equally true of a healthy starter and of someone who is retired or unsigned. Never conclude from an empty result that a player is (or is not) active, and never fill that gap from your own knowledge: check get_ecr with ranking_type="ROS" (position DL/LB/DB/IDP for defenders) and look for recent news (player_news here; getContentFP where that tool is available) before saying anything about whether a player is playing. resolved_players reports which player each name became, with their team -- check it before you answer, because NFL players do share names (Justin Jefferson is a WR in MIN and a LB in CLE). If ambiguous_names is present, one name matched several players and everything reported covers ALL of them: work out which one the user means from their roster and league type, say which one you answered about, and re-run with that player's player_id to get results for them alone.

league_analyzer

ChatGPT
Analyze league power rankings and playoff odds (NFL) or category rankings (MLB). NFL ranks and scores are horizon-dependent and the response says which horizon it used, in horizon, along with total_teams. Always state the horizon when quoting a rank or score, and never treat a team rank on one horizon as contradicting player rankings on another — a roster is legitimately ranked differently for rest-of-season than for dynasty. Every horizon is reachable on every league. The site shows Draft, Dynasty and Weekly as tabs regardless of league type, so you can always fetch the one the user is looking at: never tell a user you cannot access the dynasty (or weekly, or draft) view. If their number disagrees with yours, they are most likely reading a different tab — ask which, or re-fetch that horizon.

matchup_analyzer

ChatGPT
Get current matchup status and win probability for a league (NFL). In an elimination (guillotine) league there is no opponent: the response has league_format="elimination" and gives every alive team's chance of being safe this week instead of a win probability.

mlb_waiver_finder

ChatGPT
Find suggested MLB waiver moves, or analyze a specific potential add. Premium feature (any paid tier). On a plan without it, the response returns status='premium_required' instead of results. If so, tell the user this feature is not included in their current plan.

rank_trade_partners

ChatGPT
Rank the other teams in an NFL league by how well they fit a specific trade. Give the position the user would trade away, the position they want, or both. Returns partners best fit first, with the ranks behind each score. Call this before naming a trade partner: a viable partner is weak where the user is selling and strong where the user is buying, and the comparison is done here so it does not have to be reasoned out.

resync_league

ChatGPT
Force an upstream roster resync for a league (completes inline). Use whenever the synced roster may be stale — e.g. the user says they just added or dropped a player on their league host, a waiver claim has since been processed, or a roster read disagrees with what the user reports. start_sit_assistant already resyncs automatically before acting; other tools read the roster as last synced.

start_sit_assistant

ChatGPT
Get start/sit lineup recommendations for a league. NFL uses the FantasyPros start/sit advice endpoint. MLB uses the optimal lineup endpoint (getLeagueLineupJSON) and takes the first optimal hitter + pitcher lineups, diffed against the current roster. Only players already on the user's roster are evaluated. Free agents and other teams' players never appear, and a rostered player's expert-start percentage is not a comparison against them — to weigh starting a player the user does not roster, use get_player_comparisons instead. NFL advice is always for the league's current matchup week, reported as nfl_week. It is not advice for a later week: for "next week" or a named later week, use get_projections (weekly, that week, the roster's player_ids) and that week's byes instead. Returns status='league_not_found' when the league key is not recognized upstream, and status='no_data' when the league has no synced roster data — never treat those as a real (empty) lineup. Premium feature (any paid tier). On a plan without it, the response returns status='premium_required' instead of results. If so, tell the user this feature is not included in their current plan.

trade_analyzer

ChatGPT
Analyze a specific fantasy trade (NFL or MLB). Evaluate the trade by dynasty power-ranking change in dynasty leagues and rest-of-season change in all other leagues (the NFL response's default_* fields already pick the right one); weekly change is context only, never the primary metric. Premium feature (any paid tier). On a plan without it, the response returns status='premium_required' instead of results. If so, tell the user this feature is not included in their current plan.

trade_finder

ChatGPT
Get top trade suggestions from FantasyPros Trade Finder (NFL). All filters combine: give-side (player_trade_away, position_trade_away) and get-side (player_trade_for, position_trade_for) narrow to their intersection, so "send a TE, get an RB" is one call with both set. The response echoes filters_applied; if it comes back with an error saying nothing matched, that is a real empty result — no filter was silently dropped, so do not re-run with fewer filters and present the wider result as though it answered the question. Suggested gains are rest-of-season power-ranking changes. Evaluate trades by dynasty power-ranking change in dynasty leagues (run trade_analyzer on top candidates to get it) and rest-of-season change in all other leagues; weekly change is context only, never the primary metric. The league is resynced and trades are computed from current rosters on every call, so there is no need to call resync_league first. Every finder_ value (gains and before/after scores) comes from Trade Finder's own computation and can differ from the site's Trade Analyzer. When quoting exact power-ranking numbers to the user, run trade_analyzer on that trade and quote its numbers — for rest-of-season as well as dynasty. Premium feature (any paid tier). On a plan without it, the response returns status='premium_required' instead of results. If so, tell the user this feature is not included in their current plan.

waiver_analyzer

ChatGPT
Analyze a specific waiver move. Premium feature (any paid tier). On a plan without it, the response returns status='premium_required' instead of results. If so, tell the user this feature is not included in their current plan.

waiver_finder

ChatGPT
Find top waiver recommendations for a league (NFL). data_source names which upstream answered. With "getWaiverPlan" (the normal path) each recommended add carries priority_value — the upstream recommendation strength used for ordering: in FAAB leagues it is the suggested bid in dollars (mirrored as "FAB Bid"), otherwise a relative waiver priority. With "waiverFinder" the league's host does not support the automated waiver plan, so the response is a ranked candidate list instead: rows carry recommendation_score and expert_agreement, and there is NO priority_value and NO "FAB Bid" because bid and pending-claim data cannot be retrieved for that host. A limitations list spells this out — relay it, and never quote or estimate a bid on that path. Premium feature (any paid tier). On a plan without it, the response returns status='premium_required' instead of results. If so, tell the user this feature is not included in their current plan.

Capabilities

Writes

Example Prompts

Click any prompt to copy it.

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/fantasypros-fantasy-advice"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/fantasypros-fantasy-advice)