# NeuroRank

> NeuroRank is an esports cognitive-combine platform. Players complete a battery of timed cognitive tasks (reaction, aim, tracking, working memory, decision quality, composure, tilt resistance) and receive a scored profile across six cognitive dimensions, a player archetype, and a written report. This file tells AI agents what NeuroRank is and where its machine-readable, read-only public data lives.

## What an agent can do here

NeuroRank exposes a small, read-only public surface describing the aggregate cognitive-combine cohort. No authentication is required for the endpoints listed below, and they return only aggregate, non-personal data. A machine-readable description of this surface is published as an OpenAPI document.

- [OpenAPI description](https://neurorank.app/openapi.yaml): Machine-readable spec for NeuroRank's public, read-only endpoints.
- [Cohort summary (JSON)](https://neurorank.app/api/cohort/summary): Aggregate stats across all completed combine runs — total runs, estimated trials, median run age, and measured reliability (split-half internal consistency; a test-retest coefficient is not published yet while the repeat-run sample accrues). Updated continuously.
- [Pricing (markdown)](https://neurorank.app/pricing.md): Structured plan and price data for agents. Human version at https://neurorank.app/pricing.

## Key pages

- [Science and methodology](https://neurorank.app/science): How the combine measures, published split-half reliability, norms versioning. NeuroRank publishes measured stats only; when a sample is too small the stat is withheld rather than estimated.
- [Archetypes](https://neurorank.app/archetypes): The 18 player archetypes and how classification works.
- [Blog](https://neurorank.app/blog): Explainers on reaction time, tilt, composure, working memory, and cognitive testing for competitive gaming.
- [Leaderboard](https://neurorank.app/leaderboard): Public leaderboard of completed profiles.

## MCP endpoint (for agents)

NeuroRank serves a Model Context Protocol (MCP) endpoint over Streamable HTTP at `https://neurorank.app/mcp`. It is stateless (POST JSON-RPC; no session required) and exposes:

- `get_cohort_stats` (free) — returns the aggregate cohort summary above (runs, estimated trials, median run age, measured split-half reliability, repeat-run pairs collected). No authentication, aggregate data only.
- `get_player_report` (paid, scout or team) — returns a full cognitive-combine report for one player by profile id (NR-XXXXXX): dimension scores, archetype, written report, and coaching. Requires an API key in the `x-api-key` header, on an active scout or team subscription, limited to players who have opted in to scouting. Subject to a monthly quota.
- `get_team_report` (paid, team tier) — analyses a roster of up to 10 players and returns a composition report: per-dimension team averages, strengths and weaknesses, standouts, score spread, archetype mix, and a written team intelligence summary. Requires a team-tier API key; only opted-in players are included (minimum 2). Costs one monthly pull.

Mint an API key from your NeuroRank account (Billing > API access).

## Cognitive dimensions

NeuroRank scores each player 0–100 on six dimensions: raw speed, decision quality, composure index, tilt resistance, working memory, and consistency. Players are also assigned an archetype (e.g. "The Anchor") summarising their profile.

## Coming soon (for agents)

Per-call metered billing may follow; today, paid access is included with the scout (per-player) and team (per-player + team report) subscriptions.

## Contact

- Web: https://neurorank.app
- Operator: Loghan Lewis, professional rugby league prop (Super League) and solo builder; NeuroRank is built in public, with weekly essays at https://loghanlewis.substack.com