Generated MCP reference

Quawd MCP Tool Reference

A published map of the tools exposed by Quawd's MCP server. The catalog stays compact for scanning; exact input schemas live in linked JSON files for agents and implementers.

Endpoint

https://mcp.quawd.bot/mcp
Setup guideCatalog JSON

Tools

18

Groups

10

Schemas

18

Assess

1 tool
assess_overfitting_riskWriteUse after iterating on a strategy (multiple backtests, sweeps, walk-forward) to gauge overfitting risk. Looks at iteration trajectory, parameter sensitivity, and trade clustering. Requires at least one prior backtest.

Compare

1 tool
compare_backtestsWriteCompare statistics across multiple backtests. Useful for evaluating parameter sweeps or different strategy configurations. Pass 2-10 backtest_ids.

Explain

1 tool
explain_barWriteExplain what happened on a single bar. Returns had_entry=true with the full entry payload (trade index, prices, PnL, exit reason, per-condition state at the signal bar) if an entry fired, otherwise had_entry=false with which entry conditions failed and how close they were. Pass exactly one of bar_id or timestamp.

Get

2 tools
get_condition_timelineWriteGet condition proximity evolution over a bar range. Use to spot patterns like 'RSI was consistently near threshold but never crossed'.get_current_signal_stateWriteEvaluate a strategy's entry/exit conditions against the most recent market bar. Use before promoting to paper trading to confirm the strategy isn't already firing on stale signals.

List

2 tools
list_assetsWriteUse BEFORE creating a strategy to see what symbols, timeframes, and date ranges are available for backtesting. Returns a catalog of assets with data coverage. Call once at the start of a research session to ground subsequent builder_create calls.list_event_typesWriteUse BEFORE adding an event-driven condition (e.g. "exit before FOMC"). Returns the user's custom events plus active system events (CPI, FOMC, NFP, earnings windows).

Research projects

1 tool
project_createWriteUse at the START of a research effort to group related strategies under one thesis. Created strategies link to the project via project_id on builder_create.

Run

1 tool
run_backtestWriteRun a backtest against a compiled strategy. Workflows: single-run with `lookback` or sweep across parameter variants. Example: run_backtest(strategy_id, lookback='3mo').

Signals

1 tool
signal_scanWriteUse BEFORE building a strategy to test whether a condition has predictive power. Scan a single condition (or sweep variants) on one or more symbols and measure forward-return statistics. The signal-first workflow.

Strategy builder

7 tools
builder_add_conditionWriteAdd a condition card to the strategy. Returns card_id. Use multi-timeframe by setting 'resolution' inside lhs/rhs value refs. Use builder_get_card_schema('value_ref:indicator') for the full schema.builder_attachWriteUse after building a card to wire it into a strategy slot (entry, exit, gate, overlay). Each strategy needs at least one entry attachment to trade.builder_compileWriteUse when ready to backtest: freeze a strategy into a runnable blueprint. Call after all entries/exits/gates have been attached and sizing/risk are configured. Idempotent.builder_composeWriteUse to combine multiple existing condition cards into one logical card. Modes: allOf (AND), anyOf (OR), not, at_least (N-of-M), consecutive (N bars), sequence (ordered steps).builder_createWriteUse FIRST when starting a new strategy. Returns a strategy_id used by every subsequent builder_* call. Pick `symbols` and `timeframe` from list_assets() to ensure data coverage. `symbols` is a list — single-symbol: ['SPY']; universe: ['AAPL','MSFT','GOOG']. Example: builder_create(name='RSI mean revert', symbols=['BTC-USD'], timeframe='1h', parent_strategy_id='root').builder_get_card_schemaWriteUse BEFORE builder_add_condition to discover and inspect card types. Call with no card_type (optionally with role=entry/exit/gate) to list all available types grouped by category. Call with a specific card_type (e.g. 'trailing_stop', 'value_ref:indicator') to get its full JSON schema.builder_set_strategy_configWriteSet strategy-level config (sizing, position_policy, execution, risk) in one partial-patch call. Only keys present in patch are applied.

Walk

1 tool
walk_forwardWriteValidate a strategy's robustness via walk-forward analysis. Detects overfitting via the walk-forward efficiency (WFE) metric.
Quawd

A subscription SaaS platform for designing, backtesting, and paper-trading algorithmic trading strategies on equities and crypto — described in plain English to an AI agent, no code required.

© 2026 Quawd. All rights reserved.

Quawd is a software tool, not a broker-dealer or investment adviser, and does not provide investment advice. Trading involves substantial risk of loss. Backtested and hypothetical results have inherent limitations and are not indicative of future performance.