NeuroTrade Signal API

com.a3eecosystem.neurotrade/signal

Documentation

NeuroTrade Signal API: MCP Server

An MCP (Model Context Protocol) server exposing NeuroTrade Signal API as native tools for AI coding assistants. Drop it into Claude Code, Cursor, Zed, Continue, or any MCP-compatible client and your agent can generate AI-reasoned crypto trading signals inline.

What you get

Directional call per signal: OPEN_LONG / OPEN_SHORT / CLOSE
Confidence score (0.0 to 1.0) with calibrated uncertainty
Entry / Take-Profit / Stop-Loss prices with computed R:R ratio
Natural-language thesis explaining the setup in plain English
Technical breakdown: RSI, MACD, EMA-9/21, ATR, regime tag
Risk flags: news spike aftermath, RSI extremes, low liquidity, etc.
Quota visibility via get_quota tool

Supported markets

25 pairs across three tiers (majors, high-cap, volatile movers): BTC, ETH, SOL, XRP, DOGE, BNB, TAO, TRX, WLD, FET, SUI, PEPE, ADA, LTC, LINK, BCH, ENA, AVAX, DOT, AAVE, NEAR, TON, UNI, APT, ARB.

Timeframes: 1m, 5m, 15m, 30m, 1h, 4h, 1d.

Install

1. Get an API key

Pick a plan on one of the marketplaces:

Zyla: https://zylalabs.com/api-marketplace/other/neurotrade+signal+api/12432
RapidAPI (legacy, still active): https://rapidapi.com/cooa3e/api/neurotrade-signal

Free tier: 10 calls/month. Paid plans from $24.99/mo.

2. Set the environment variable

export NEUROTRADE_API_KEY=nt_your_key_here

3. Install the package

pip install git+https://github.com/A3E-ECOSYSTEM/neurotrade-signal-api-mcp.git

Or clone + install for local dev:

git clone https://github.com/A3E-ECOSYSTEM/neurotrade-signal-api-mcp.git
cd neurotrade-signal-api-mcp
pip install -e .

4. Wire into your MCP client

Claude Code (~/.claude/mcp.json or project .mcp.json):

{
  "mcpServers": {
    "neurotrade-signal": {
      "command": "python",
      "args": ["-m", "mcp_server.neurotrade_mcp"],
      "env": {
        "NEUROTRADE_API_KEY": "nt_your_key_here"
      }
    }
  }
}

Cursor: same JSON in your Cursor MCP config. Zed: add to ~/.config/zed/settings.json under context_servers. Continue: add to ~/.continue/config.json under contextProviders.

Restart your editor. The tools appear automatically.

Tools exposed

`generate_signal(symbol, timeframe, strategy, personality, include_thesis)`

Generates an AI-reasoned signal. Returns JSON with direction, confidence, entry, TP, SL, thesis, reasoning, technical breakdown, and risk flags.

Parameters:

symbol (required): trading pair, e.g. "BTC/USDT"
timeframe (required): "1m", "5m", "15m", "30m", "1h", "4h", "1d"
strategy (optional): "trend_rider", "mean_reversion", "breakout", "momentum"
personality (optional): "scalper", "swing", "position"
include_thesis (optional, default true): include LLM thesis + reasoning

`get_quota()`

Returns your current plan, calls used, calls remaining, and reset timestamp. No arguments.

`list_supported_symbols()`

Returns the 25-pair catalog grouped by tier.

`list_supported_timeframes()`

Returns the 7 supported timeframes.

Run standalone

NEUROTRADE_API_KEY=nt_xxx python -m mcp_server.neurotrade_mcp

Useful for debugging or as a long-running service.

License

MIT. See LICENSE.