analyze_coinLABSCO SUMMARY
A chart-reading session that ends with a number you can rank on.
Instead of eyeballing indicators one at a time, you get them computed together and compressed into a single 0–100 score — which means the same question can be asked of fifty coins at once and answered in an order.
What it is
An MCP server for crypto technical analysis. It pulls a long daily window of OHLCV candles from Binance, falling back to CoinGecko, and computes an indicator suite plus a 0–100 Trend Strength Score on top of them.
What you get
- A full analysis of one coin: moving-average alignment across short, medium and long windows, RSI, MACD, ADX, ATR, Bollinger Band width and annualised historical volatility
- RSI divergence over the last 30 bars, On-Balance Volume direction, and the current USDT-M perpetual funding rate
- A volatility regime — low, mid or high — classified by a 3-state GaussianHMM, with a rule-based fallback
- A Trend Strength Score from 0 to 100 — trend, momentum and strength against BTC weighted together, then adjusted by 4H alignment and OBV
- Any list of coins ranked by that score, computed in parallel across several workers
- The top 50 coins by market cap ranked the same way, with stablecoins, wrapped tokens and anything under $10M daily volume filtered out
- A 30-day return correlation matrix across a set of coins, with tight clusters above 0.85 and the uncorrelated pairs called out
- Macro framing from CoinGecko global data and DefiLlama stablecoin supply
- A watchlist you can add to, show and clear down
Requirements
Ask for a coin by symbol: candles come from Binance, so anything without a Binance listing takes the CoinGecko fallback. Market-cap ranking and macro context come from CoinGecko, stablecoin supply from DefiLlama — those three services are what the answers depend on being up.
Setup effort
Build from source — git clone https://github.com/cryptographer11/cryptoscholar.git
USE CASES
01
Deciding whether one coin is set up
Run the full snapshot — trend, momentum, volatility, divergence, funding — and read the Trend Strength Score it lands on.
02
Picking from a shortlist you already have
Pass the symbols and get them back ranked by score, with regime and RSI divergence alongside.
03
Starting from nothing
Take the top 50 by market cap ranked by score, with stablecoins, wrapped tokens and thin-volume names already filtered out.
04
Checking a portfolio is not one bet in five costumes
Compute the 30-day correlation matrix and read the clusters above 0.85.
05
Framing a coin against the market it trades in
Read the macro signals before deciding whether a score means anything this week.
06
Watching the same names each morning
Add the symbols to the watchlist, show it, and drop the ones that have stopped being interesting.
15 tools
Listed in project docs. Your client may show a different list.
rank_coinstop_coinscorrelate_coinsmarket_contextdebatewatchlist_addwatchlist_removewatchlist_showwatchlist_listswatchlist_scanalert_setalert_checktrain_regime_modelgenerate_reportBEFORE YOU INSTALL
What to know first
Tools you get
15 tools named in the project's documentation.
Requires
Python 3.11+
Optional
ANTHROPIC_API_KEY in your environment
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
TERMINAL
git clone https://github.com/cryptographer11/cryptoscholar.git
cd cryptoscholar
make install
cp .env.example .envMCP Clients
Claude Code
CLAUDE CODE · JSON
{
"mcpServers": {
"cryptoscholar": {
"command": "<YOUR_PATH_TO_CRYPTOSCHOLAR>/venv/bin/python",
"args": [
"-m",
"cryptoscholar"
],
"env": {
"ANTHROPIC_API_KEY": "<YOUR_ANTHROPIC_API_KEY>"
}
}
}
}<YOUR_PATH_TO_CRYPTOSCHOLAR> is the folder you cloned. make install puts CryptoScholar into that folder's venv, so the client has to start that venv's Python — a bare python will not find the package. ANTHROPIC_API_KEY is only used by the debate and generate_report tools; leave it out and the other 13 still work.
