Skip to content

Tools

The Reversion MCP server exposes 46 tools to AI agents. Each tool maps to a backend API endpoint and is gated by token scopes (read, trade, or algo).

Tools are auto-discovered by MCP clients — your agent sees names, descriptions, and parameter schemas without any manual configuration.

Scope: read

ToolDescriptionKey Parameters
get_account_overviewAccount snapshot: balance, positions, open orders, TWAPs, recent fills, algo runs, wallet statusexchangeId
get_watchlistGet the user’s watchlist of tracked trading pairs
add_to_watchlistAdd a trading pair to the watchlistsymbol, exchangeId
remove_from_watchlistRemove a trading pair from the watchlistsymbol, exchangeId

Scope: read

ToolDescriptionKey Parameters
get_tickerCurrent ticker: price, bid, ask, 24h volume and changeexchangeId, symbol
search_marketsSearch available trading pairs with prices and volumeexchangeId, search
get_orderbookOrderbook bids and asks for a symbolexchangeId, symbol, depth
get_candlesOHLCV candlestick dataexchangeId, symbol, timeframe, limit
get_funding_rateCurrent funding rate for a perpetual futures symbolexchangeId, symbol
get_mark_priceMark price used for PnL and liquidation calculationsexchangeId, symbol
get_open_interestCurrent open interest for a symbolexchangeId, symbol

Scope: trade

ToolDescriptionKey Parameters
place_market_orderMarket order for immediate fill; use reduceOnly to close positionssymbol, side, amount, reduceOnly, exchangeId
place_limit_orderLimit order at a specific pricesymbol, side, amount, price, reduceOnly, postOnly, exchangeId
place_stop_lossStop-loss trigger order (market or limit execution)symbol, side, amount, triggerPrice, executionType, limitPrice, exchangeId
place_take_profitTake-profit trigger order (market or limit execution)symbol, side, amount, triggerPrice, executionType, limitPrice, exchangeId
place_twap_orderTWAP order — splits execution over a duration to reduce market impactsymbol, side, size, durationMinutes, reduceOnly, exchangeId
cancel_orderCancel a specific open order by IDsymbol, orderId, exchangeId
cancel_all_ordersCancel all open orders, optionally filtered by symbolsymbol, exchangeId
cancel_twap_orderCancel an active TWAP ordertwapId, symbol, exchangeId

Scope: trade

ToolDescriptionKey Parameters
set_leverageSet leverage multiplier for a symbolsymbol, leverage, exchangeId
set_margin_modeSet margin mode: cross (shared) or isolated (per-position)symbol, marginMode, exchangeId
close_positionClose a position via reduce-only market order (auto-fetches size)symbol, exchangeId

Scope: algo

ToolDescriptionKey Parameters
generate_chartCandlestick chart (PNG) with optional indicators and overlayssymbol, exchange, timeframe, limit, indicators, overlays, show_volume
explore_indicatorTest an indicator + trigger condition against real price data before creating an algorithmsymbol, exchange, timeframe, indicator_name, indicator_params, signal_type, threshold
fetch_candles_jsonRaw OHLCV candle data as columnar JSON (no chart)symbol, exchange, timeframe, limit

Scope: algo

ToolDescriptionKey Parameters
list_indicatorsList available indicator types for entry/exit conditions
get_indicator_schemaGet parameter schema for a specific indicator typeindicatorType
list_algorithmsList all user algorithm configs with names, IDs, symbols, status
get_algorithmGet full algorithm config: entry/exit rules, indicators, position sizingalgoId
create_algorithmCreate a new algorithm config with nested indicator and rule definitionsalgoName, params
update_algorithmUpdate an algorithm config (creates new version); supports partial updatesalgoId, algoName, params

Scope: algo

ToolDescriptionKey Parameters
run_backtestRun a backtest over a time range; returns a run ID to pollalgoId, version, coinSymbol, exchangeId, startTime, endTime, capitalScaler
get_backtest_resultsGet metrics for a completed backtest: config, swap metrics, algo metricsrunId
plot_backtestGenerate equity curve + drawdown chart (PNG) for a backtestrunId, start, end, include_json, max_points
rolling_cv_backtestRolling forward-walk cross-validation backtest to prevent overfittingalgo_id, version, symbol, exchange_id, start_time, end_time, train_window_days, test_window_days, step_days
list_runsList algo runs (backtests and live) with statusalgoId, status
update_run_statusUpdate run status (e.g., STOPPED, CANCELLED)runId, status

Scope: algo

ToolDescriptionKey Parameters
bayesian_optimizeHands-off Bayesian optimization (Optuna TPE); runs full loop server-sidealgo_id, version, symbol, exchange_id, start_time, end_time, n_trials, param_space, param_mapping, scoring_code
get_optimization_statusPoll optimization job status, progress, best score/params, trial historyjob_id
stop_optimizationEarly-stop a running optimization; results so far are preservedjob_id
bo_suggestAgent-driven BO step: given past observations, returns next suggested paramsparams_observed, scores_observed, constraints, batch_size, direction
compute_backtest_scoreRun custom Python scoring function on backtest results in sandboxresults, code
plot_optimizationGenerate custom matplotlib plot of optimization results (PNG)trajectory, results, scores, code

Scope: read

ToolDescriptionKey Parameters
get_economic_calendarQuery upcoming economic events filtered by date, impact, or currencyfrom, to, impact, currency, limit
get_upcoming_eventsEconomic events happening in the next N hourshours
get_high_impact_eventsHigh-impact events (FOMC, NFP, CPI) for the next N dayscurrency, days
get_market_hoursTrading hours for a specific exchange or all exchangesexchange
is_market_openCheck if a specific exchange is currently openexchange