SuperTruth · Property Taxonomy
Intelligence Glass Box
0 live sources · 23,091 entities · 0 agent findings · refreshed 6:18:24 PM
Sports
140
Leagues
528
Teams
703
Athletes
21,720
Avg DTI
0
Fresh
0
Aging
0
Stale
0
Findings
0
Field-Based Growth · Last 14 Days
Taxonomy Intelligence Timeline
Each line normalized to its own scale.
Shape = growth rate. End label = actual count.
Sports Intelligence Memory
NEWThis is what separates LEO from every other sponsorship tool. Instead of matching keywords, LEO now learns from every deal, every social engagement pattern, and every DTI score — then synthesizes structured intelligence: which brand categories have proven affinity with each sport, how audiences overlap across properties, what activation channels deliver best, and which sports form coherent sponsorship portfolio clusters. When you classify "Red Bull Rampage 2026", LEO knows it belongs to the Action Sports family and that F1, Surfing, and Snowboarding share the same brand universe — not because it was told, but because it learned it from the data.
—
Run synthesis to activate
Watches the taxonomy for MISSING data first, staleness second. Sources: Wikipedia infoboxes (venue, coach, founded, colors, description) + social media bios. AI diffs the current web against stored findings and proposes candidates — nothing committed without human review.
Endpoint
/api/mcp/feed-monitorSchedule
Daily 03:00 UTC
Tools
6
get_stale_entitiesfetch_wikipediadiff_wikipediaread_social_biopropose_updateget_pending_proposalsWatches ESPN's public API across 27 leagues for sponsorship-relevant changes: transactions (trades/signings/waivers), injuries, news headlines with sponsor-signal detection, and roster gaps — players ESPN lists that our taxonomy doesn't yet have.
Endpoint
/api/mcp/espn-feedSchedule
Daily 04:00 UTC
Tools
7
get_live_scoresget_team_newsget_transactionsget_injury_reportdetect_roster_changespropose_updateget_pending_proposalsWatches every stored Twitter/X handle for changes that impact sponsorship value: follower count shifts >5% flagged (>20% = breakout opportunity signal), deactivated or suspended accounts, bio description changes, and sponsorship language in bios — "powered by", "official partner", "#ad". Runs across all 346 verified handles in the taxonomy.
Endpoint
/api/mcp/social-monitorSchedule
Daily 05:00 UTC
Tools
7
detect_follower_changesdetect_handle_changesdetect_bio_changesget_trending_entitiesdetect_sponsorship_signalspropose_updateget_pending_proposalsProvides activation-ready weather forecasts and risk scoring for every venue in the taxonomy. Source: Open-Meteo (free, no key required) + Nominatim geocoding for venues without stored lat/lon. Rain at an open-air stadium suppresses attendance 15–25%. Extreme heat or cold changes dwell time and sponsor visibility windows. Scores each venue's activation risk as low/moderate/high/severe for any target date.
Endpoint
/api/mcp/weatherSchedule
Daily 06:00 UTC
Tools
4
get_venue_forecastget_activation_risk_scoreget_historical_weatherget_venues_by_weather_riskDetects events near venues that compete for audience attention or create activation conflicts. Source: SeatGeek public API (free, no key) for concerts/sports/theater; PredictHQ (key optional) adds political rallies, conferences, disasters, school breaks. A Taylor Swift tour date in-market during a playoff game changes parking, hotels, and media saturation. Scores 7-day conflict risk for all venues.
Endpoint
/api/mcp/local-eventsSchedule
Daily 06:30 UTC
Tools
4
get_events_near_venuedetect_scheduling_conflictsget_major_events_by_cityassess_attendance_riskSurfaces demographic and economic context for venue catchment areas — the population that actually attends events and sees brand activations. Source: DataSpine (LEO's own household intelligence API — Census ACS, CDC PLACES health, EPA environment, HUD housing, FCC broadband). Computes a brand activation score (0–100) per venue market and scores brand-demo fit against configurable target profiles.
Endpoint
/api/mcp/market-intelSchedule
Daily 07:00 UTC
Tools
4
get_venue_market_profilescore_brand_fitcompare_venue_marketsget_top_markets_by_attributeConverts broadcast reach + sponsor visibility into CPM-equivalent activation value for every property. Uses Nielsen/MRI-Simmons public benchmarks for 20 networks (ESPN, NBC, Fox, Amazon, etc.) to estimate impressions per game. Scores fair market value ranges for 8 placement types: naming rights, jersey patch/sponsor, broadcast lower-third, stadium boards, digital overlay, title sponsor. Inscape-ready: when Snowflake is connected, replaces estimates with actual second-by-second viewership.
Endpoint
/api/mcp/media-valueSchedule
Daily 07:30 UTC
Tools
4
get_broadcast_reachget_media_value_by_propertyget_top_properties_by_reachget_network_scheduleMonitors the global sponsorship landscape for competitive intelligence. Source: BusinessWire, PRNewswire, and GlobeNewswire RSS feeds (all free, no key) scanned daily for naming rights deals, jersey sponsorships, title deals, and partnership announcements. Identifies which brand categories are active in each sport, flags whitespace (categories with no sponsor in a given property), and maps the competitive landscape per league.
Endpoint
/api/mcp/brand-intelSchedule
Daily 08:00 UTC
Tools
5
scan_sponsor_announcementsdetect_naming_rights_dealsget_brand_activity_by_sportfind_whitespaceget_competitive_landscapeThe intelligence memory of LEO. Reads every sponsorship deal, social engagement pattern, and DTI score across the full taxonomy, then uses AI to synthesize structured sponsorship knowledge per sport: brand affinities, audience demographics, activation patterns, comparable properties, and cross-sport family relationships. Powers the recommendation engine — "Red Bull Rampage" now surfaces Action Sports properties, not generic rugby.
Endpoint
/api/mcp/sports-knowledgeSchedule
Weekly (10 sports/run)
Tools
6
get_sport_knowledgequery_knowledge_for_brieffind_similar_sportsget_all_knowledgesynthesize_sportget_sponsorship_familiesAutonomous Intelligence
Agent Environment
0 active findings · ? teams with gaps · flywheel running daily
A&R proprietary. Monitors X daily for sponsorship-relevant signals across the entire taxonomy. Extracts six signal types: DEAL_SIGNAL (announcements, partnerships), BRAND_SAFETY (controversy, risk flags), BREAKOUT (trending athlete/team — emerging opportunity), SENTIMENT_SHIFT, NAMING_RIGHTS, COMPETITOR (rival brands activating in sports). An AI model synthesizes raw tweets into confidence-scored findings with evidence. Signals surface on entity detail pages alongside other intelligence.
not yet run
Dispatches 7 parallel sub-agents per team: Wikipedia summary, Wikipedia infobox, school athletics website, NCAA news feed, X/Twitter, ESPN schedule, and ESPN structured data. AI extracts structured fields — head coach, founding year, stadium, championships, mascot, roster reach. High-confidence findings are auto-promoted to DB columns. DTI re-scored after every run.
not yet run
Fetches kit colors, venue details, injury reports, and recent news for every team in ESPN-supported leagues (NFL, NBA, MLB, NHL, MLS, EPL, CFB, CBB). Colors written directly to kitColors. ESPN ID resolution works even for teams with non-ESPN external IDs via slug-based lookup.
not yet run
Franchise valuations for 51 top teams from Forbes via Zyla API 9491. Multi-year history 2016–2024. Displayed as Forbes Value stat card and year-over-year history chart on team pages. Refresh annually.
not yet run
Contract data for ~100 top-earning athletes from Zyla API 8773. Total value, AAV, duration, avg per game. Surfaced on athlete detail pages. Data quality: NFL/NBA/MLB current, some soccer entries stale.
not yet run
Backfills logoUrl for college teams via NCAA API slugs. 1,173 NCAA schools. Three-pass matching. SVG format. Run after any college team import.
not yet run
Live W/L records for CFB FBS/FCS, CBB D1, CWBB D1 from ncaa-api.henrygd.me. Shown as Season Record section on team pages.
not yet run
Flywheel — continuously scans for teams with missing fields and dispatches enrichment batches. Runs daily. "Gaps detected" panel shows exactly which teams and fields need attention.
not yet run
Programmatically parses Wikipedia infoboxes for every team — all sports, all fields, zero hardcoding. Stores in metadata.wikipedia as "Program Info" on detail pages.
not yet run
Free NFL data from nflverse GitHub releases — no key, nightly updates. Rosters (matched via espn_id/sportradar_id), depth charts, season stats. Replaces Zyla CFB/Contracts for NFL.
not yet run
Jeff Sackmann open-source ATP/WTA data since 1968. Rankings, W/L records. The most complete tennis dataset in existence. Free from GitHub.
not yet run
Free event-level soccer data (EPL, La Liga, UCL, NWSL). Competition catalog + team W/D/L/GD. Professional teams pay six figures for this; free on GitHub.
not yet run
All F1 data via api.jolpi.ca — free, no key, back to 1950. Driver standings, race schedules. Matches to existing F1 athletes by name.
not yet run
NBA (hoopR) + CFB (cfbfastR) data via GitHub releases. Team W/L, PPG, RPG, offensive/defensive rankings. Same free pattern as NFLverse.
not yet run
Continuously searches GitHub for sports data goldmines. 20 queries across all sports. Scores by stars/freshness/format. Logs discoveries to AgentFinding. Already mapped: nflverse, Sackmann, StatsBomb, hoopR, cfbfastR, Ergast.
not yet run
A&R proprietary. SerpAPI Google Trends (any sport globally) + DataSpine home market demographics (Census ACS) + AI synthesis → AudienceProfile records (age, income, geo, interests, digital behavior).
not yet run
Always-on watchdog — runs after every sync. Three layers: roster contamination, pro team name corruption (vs ESPN), + 9 extended checks (duplicates, orphans, externalId collisions, career overlaps, name formats, stale DTI, league-sport mismatches).
not yet run
Pre-presentation QA sweep — checks logo contamination (pro logos on college teams), ghost links, duplicate team names, empty slugs, athletes active on 3+ teams simultaneously. Writes critical findings to the review queue. Run before any client demo.
not yet run
Data Providers
Source Registry
Configured sources
Free + agent sources