Is binance-intelligence-mcp safe to install?

SaferSkills independently audited binance-intelligence-mcp (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.

CategoryWeightCategory scoreContribution

Securityprompt, exec, net, exfil, eval

35%

100

35.0 pts

Supply chainhash, typosquat, maintainer, lockfile

20%

100

20.0 pts

Maintenancestaleness, pinning, CI

15%

100

15.0 pts

TransparencySKILL.md, perms, README

15%

100

15.0 pts

Communityinstalls, verify, response

15%

100

15.0 pts

Findings & checks · 0 flagged

Securityscore 100 · 0 findings

✓—All security checks passedNo findings in this category for the latest scan.pass

Supply chainscore 100 · 0 findings

✓—All supply chain checks passedNo findings in this category for the latest scan.pass

Maintenancescore 100 · 0 findings

✓—All maintenance checks passedNo findings in this category for the latest scan.pass

Transparencyscore 100 · 0 findings

✓—All transparency checks passedNo findings in this category for the latest scan.pass

Communityscore 100 · 0 findings

✓—All community checks passedNo findings in this category for the latest scan.pass

The primary manifest — the file an agent reads to learn what this artifact does.

SKILL.md · 1 file

SKILL.md6.7 KB · Markdown

name: binance-intelligence-mcp

description: |

MCP server providing 8 computed intelligence tools for Binance futures markets.

Unlike raw API wrappers, each tool combines multiple public endpoints into

derived analytics: accumulation detection, whale tracking, market impact

simulation, smart money radar, pattern scanning, correlation matrix,

regime classification, and DCA backtesting. Install via pip, no API keys needed.

metadata:

version: "1.0.0"

author: mefai-dev

license: MIT

binance-intelligence-mcp

A Python MCP server that provides 8 computed intelligence tools for Binance futures markets. Each tool combines 2-6 Binance API endpoints into derived analytics with composite scoring algorithms ··· not just raw data forwarding.

Why This Is Different

Existing Binance MCP servers forward raw API responses. This package computes derived intelligence:

Accumulation Detector: 4-factor composite score from volume, OI, funding, and taker ratios
Whale Scanner: Trade classification by tier ($50K/$250K/$1M) with net pressure analysis
Impact Simulator: Order book walk computing slippage, levels consumed, average fill
Smart Money Radar: 6-factor weighted score from positioning, sentiment, and momentum
Pattern Scanner: Candlestick pattern recognition with confidence scoring
Correlation Matrix: Pearson correlation coefficients across trading pairs
Regime Classifier: ADX/ATR-based market regime labeling
DCA Backtester: Dollar-cost averaging vs lump-sum historical comparison

Installation

pip install binance-intelligence-mcp

Usage

As MCP Server

binance-intelligence-mcp

MCP Client Configuration

{
  "mcpServers": {
    "binance-intelligence": {
      "command": "binance-intelligence-mcp"
    }
  }
}

Tools Reference

`detect_accumulation`

Combines klines, open interest history, premium index, and taker buy/sell ratio to detect smart accumulation.

Parameters:

Name	Type	Default	Description
`symbols`	list[str]	top 12	Trading pairs to analyze

Output: Per-symbol scores (volume_surge, oi_buildup, stealth_mode, buyer_aggression) and composite 0-100.

`scan_whale_trades`

Scans aggregate trades for large orders, classifies by tier, computes net pressure.

Parameters:

Name	Type	Default	Description
`symbols`	list[str]	top 6	Trading pairs to scan
`min_usd`	float	50000	Minimum trade size in USD

Output: Classified trades (Dolphin/Whale/Mega), net buy/sell pressure, biggest trade.

`simulate_market_impact`

Walks the order book to simulate large order execution impact.

Parameters:

Name	Type	Default	Description
`symbol`	str	BTCUSDT	Trading pair
`side`	str	BUY	BUY or SELL
`amount_usd`	float	100000	Order size in USD

Output: Levels consumed, avg/worst fill, slippage %, impact rating.

`smart_money_radar`

6-factor composite analysis of smart money positioning.

Parameters:

Name	Type	Default	Description
`symbols`	list[str]	top 12	Trading pairs to analyze

Output: Per-factor scores (-1 to +1), composite 0-100, bias label (BULLISH/BEARISH/NEUTRAL).

`scan_candlestick_patterns`

Detects classic candlestick patterns with confidence scoring.

Parameters:

Name	Type	Default	Description
`symbols`	list[str]	top 12	Trading pairs
`interval`	str	4h	Candlestick interval (1h/4h)

Output: Detected patterns (Hammer, Engulfing, Doji, Morning Star, etc.) with confidence 0-100.

`compute_correlation_matrix`

Computes Pearson correlation coefficients between trading pair prices.

Parameters:

Name	Type	Default	Description
`symbols`	list[str]	top 8	2-20 trading pairs
`interval`	str	4h	Candlestick interval
`limit`	int	90	Lookback periods

Output: Full correlation matrix, strongly correlated pairs (|r| >= 0.7).

`classify_market_regime`

Classifies market regime using ADX, ATR, and volume analysis.

Parameters:

Name	Type	Default	Description
`symbols`	list[str]	top 12	Trading pairs

Output: Regime label (TRENDING/RANGING/VOLATILE_BREAKOUT/LOW_ACTIVITY), direction, ADX/ATR values.

`backtest_dca`

Backtests DCA strategy vs lump-sum investing over historical data.

Parameters:

Name	Type	Default	Description
`symbol`	str	BTCUSDT	Trading pair
`amount_per_interval`	float	100	USD per purchase
`interval_days`	int	7	Days between purchases
`total_days`	int	365	Historical lookback

Output: DCA vs lump-sum ROI comparison with winner declaration.

Binance Endpoints

All endpoints are public (no API key required):

/fapi/v1/klines ··· Candlestick/kline data
/fapi/v1/aggTrades ··· Aggregate trades
/fapi/v1/depth ··· Order book
/fapi/v1/premiumIndex ··· Funding rate
/futures/data/openInterestHist ··· Open interest history
/futures/data/topLongShortPositionRatio ··· Top trader positions
/futures/data/topLongShortAccountRatio ··· Top trader accounts
/futures/data/globalLongShortAccountRatio ··· Global sentiment
/futures/data/takerlongshortRatio ··· Taker buy/sell volume

Testing

pytest tests/ -v

142 mock-based tests, zero API keys needed. Tests cover all 8 tools, the API client, helper functions, edge cases, and error handling.

Architecture

MCP Client ·�� stdio ·�� server.py (FastMCP, 8 tools)
                        ·�� tools/*.py (scoring algorithms)
                          ·�� client.py (async aiohttp)
                            ·�� Binance Public API

Each tool module is a pure async function: (client, params) ·�� dict. The server module wraps these with @mcp.tool() decorators and manages client lifecycle.

Use Cases

Pre-trade analysis: Run accumulation detector + smart money radar before entering a position
Whale monitoring: Track large trade flow to identify institutional activity
Risk assessment: Simulate market impact before placing large orders
Portfolio diversification: Use correlation matrix to avoid over-correlated positions
Strategy backtesting: Compare DCA vs lump-sum for any trading pair
Market context: Classify current regime to select appropriate trading strategy

License

MIT

binance-intelligence-mcp— agent skill

Is binance-intelligence-mcp safe to install?

binance-intelligence-mcp

Why This Is Different

Installation

Usage

As MCP Server

MCP Client Configuration

Tools Reference

`detect_accumulation`

`scan_whale_trades`

`simulate_market_impact`

`smart_money_radar`

`scan_candlestick_patterns`

`compute_correlation_matrix`

`classify_market_regime`

`backtest_dca`

Binance Endpoints

Testing

Architecture

Use Cases

License

Audit the pieces. Scan the whole. Decide.