Loading ToolBench

Reading the current corpus…

mcp-binance-futures MCP server · C · ToolBench

mcp-binance-futures

MCP server for Binance USDT-M Futures trading. Tools for market data, account balances, open positions, order placement/modification/cancellation, leverage and margin-type control.

2026-07-28+STDIO6 tools

63/100

Definition Quality80

Protocol Readiness45

Supportability45

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence

Current-spec patterns detected

Summary

Server has 6 read-only market data tools with clear naming conventions (verb-noun pattern). All tools have descriptions and input schemas with type definitions. However, descriptions lack depth on WHEN to use each tool and don't explain dependencies or common workflows. Parameter descriptions are present but minimal, most are 1-2 sentences. Output schemas are not explicitly documented in source code. Error handling exists via ErrorHandlingMiddleware but responses do not guide recovery. Tools are narrowly scoped (one symbol at a time) which is good for composition, but no batch variants offered. Overall: functional but token-wasteful, descriptions could be 2-3x richer, and output schemas should be documented.

Tools (6)

get_klinesread onlysource verified85/100

Get OHLCV candlestick data for a symbol. Returns list of dicts with: openTime, open, high, low, close, volume, closeTime, quoteVolume, trades, takerBuyVolume, takerBuyQuoteVolume.

get_order_bookread onlysource verified83/100

Get order book bids and asks for a symbol. Returns top `limit` bids and asks as [[price, qty], ...] lists.

get_recent_tradesread onlysource verified75/100

Get the most recent public trades for a symbol.

get_symbol_inforead onlysource verified77/100

Get trading rules for a symbol: tick size, lot size, min notional, max leverage, etc.

get_tickerread onlysource verified80/100

Get latest price, 24 h stats, and mark/index prices for a symbol. Returns a merged dict with: - price, priceChange, priceChangePct, high, low, volume, quoteVolume - markPrice, indexPrice, fundingRate, nextFundingTime

pingread onlyauthsource verified78/100

Test connectivity to the Binance Futures API. Returns {} on success.

Issues (5)

HIGHresponse-shaper ↗

Output schemas not documented in source code. Tool descriptions state what fields are returned but no formal JSON Schema provided for LLM downstream planning.

pingget_tickerget_order_bookget_recent_tradesget_klinesget_symbol_info

HIGHtool-description ↗

Descriptions lack context on WHEN to use each tool. 'Get latest price' does not explain when to call get_ticker vs get_klines, or whether to use mark price vs index price for different trading scenarios.

get_tickerget_recent_tradesget_klinesget_symbol_info

MEDIUMtool-description ↗

get_recent_trades description is minimal ('Get the most recent public trades for a symbol'). No explanation of what fields are included, whether data is real-time, or when this is preferred over get_klines.

get_recent_trades

MEDIUM

Recommendations

Expand get_ticker description from current text to: 'Fetch the latest price, 24h price changes, and funding info for a symbol. Use this for current market prices; use get_klines for historical data or candlesticks. Returns: price, priceChange, priceChangePct, high, low, volume, quoteVolume, markPrice, indexPrice, fundingRate, nextFundingTime.'
Document output schemas formally (as JSON Schema objects or detailed field lists) in docstrings or a separate schema file. E.g., get_ticker output: {price: string, priceChange: string, priceChangePct: string, high: string, low: string, volume: string, quoteVolume: string, markPrice: string, indexPrice: string, fundingRate: string, nextFundingTime: integer}.
Add dependency hints to descriptions. E.g., get_symbol_info: 'Call this first to validate a symbol and retrieve trading rules (lot size, tick size, leverage limits) before placing orders.'
Implement batch variants: get_tickers(symbols: [string], limit: int) to fetch data for multiple symbols in one call. This reduces token waste for common workflows.
Enhance error responses in ErrorHandlingMiddleware. When a Binance API call fails, catch the specific error code and provide a recovery hint. E.g., 'Invalid symbol: BTCXYZ. Valid symbols end in USDT (e.g., BTCUSDT) or other quote assets. Call get_symbol_info or check Binance documentation.'
Add a discovery tool (e.g., list_symbols or search_symbols) so agents can enumerate or filter trading pairs before calling market data tools. Current server assumes agents already know valid symbols.

Score history

Overall score trend

First recorded score · v2 rubric

63/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	C	63	2026-07-28+	v2