analytics-engine

Analytics Engine Skill

Purpose

The analytics engine is the orchestration brain. It does NOT analyze directly — it coordinates all other skills and agents, then fuses their outputs into a unified decision.

Trigger

Invoke when:

• Complex multi-skill analysis is needed
• Running the full prediction pipeline
• Coordinating parallel agent execution
• Post-match learning and model evolution

Orchestration Flow

User Query
    ↓
Analytics Engine (this skill)
    ↓ decomposes
    ├─→ data-provider skill    → normalized data
    ├─→ team-analysis skill    → team profiles  (parallel)
    ├─→ player-analysis skill  → player maps    (parallel)
    │
    ↓ aggregates
    ├─→ match-prediction skill → 5-model fusion
    ├─→ simulation agent       → Monte Carlo
    │
    ↓ validates
    ├─→ debate agent          → risk critique
    │
    ↓ stores
    └─→ memory agent          → evolution update
    │
    ↓
Final Analytics Report

Supported Analysis Types

Type	Skills Invoked
match_prediction	team-analysis + player-analysis + match-prediction
team_deep_dive	team-analysis + player-analysis
tournament_outlook	all teams + schedule-fetcher + match-prediction
player_comparison	player-analysis × N
post_match_learning	performance-evaluator + self-learning-optimizer

Output Schema

{
  "analysis_type": "analytics_report",
  "confidence": 0.0,
  "model_sources": [],
  "results": {
    "primary_finding": "",
    "supporting_analyses": [],
    "risk_report": {},
    "evolution_state": ""
  }
}

Rules

• Coordinate, don't duplicate. Delegate to other skills.
• Parallel execution where agents are independent.
• Always validate results through debate agent before returning.
• Update memory after every analysis pipeline.