Skip to main content

Backlog

29

Under consideration

StrategyQuant

MCP integration — expose all databank columns (including custom ones) and add a bulk stats call

The MCP server in build 144 is genuinely useful, but get_strategy_stats only returns SQX's built-in statistics. Custom databank columns created via Custom Analysis snippets aren't available, and stats can only be fetched one strategy at a time. Both limits make it impractical to use the MCP connection for real portfolio work. 1. Return custom databank columns (highest value) Many users add custom scoring columns via Custom Analysis — robustness scores, retention metrics, composite rankings. These are often the columns we actually select strategies on, but they're invisible over MCP, so any AI-assisted analysis has to fall back to manual CSV exports. Ideally get_strategy_stats would return whatever columns the databank's currently selected view displays, so custom and built-in columns come through identically and the view acts as the column filter. 2. Add a bulk endpoint Fetching stats one strategy at a time doesn't scale — a databank of 250 strategies means 250 round trips, and calls time out on larger projects. A single call such as get_databank_stats(project, databank) returning all rows at once (or paginated) would make analysis across a full databank practical. 3. Optional: expose the export function An MCP equivalent of "Export databank contents" — returning the databank as CSV — would achieve points 1 and 2 in one step and require little new logic, since the export already exists in the UI. Why this matters The current workflow for AI-assisted analysis is: manually export each databank to CSV, sync the files somewhere the assistant can read them, repeat every time anything changes. With the above, the MCP connection would replace that entirely. Given build 144 positions SQX as an AI-integrated platform, closing this gap makes the integration genuinely usable rather than a demonstration.

Rob Tabuteau
StrategyQuant

Request to StrategyQuant: MCP Integration Access for KI and clients

1) Confirmation/Setup of the Existing MCP Interface Confirm MCP server status: Does the MCP server run automatically and continuously alongside the SQX instance, or does it need to be separately activated/kept running (even when the GUI is minimized/in the background)? Clarify port reservation: The MCP server shares its port with "Remote Access" (default 8080, or the first available port). Can a fixed, dedicated port be configured so external clients don't need to search for the current port after every restart? Authentication/security: Not mentioned in the current documentation — is there (or is it planned) any access control for the MCP server (token/password), or is it open by default to any local process on localhost? Concurrent access: SQX currently only allows one running instance at a time (confirmed via CLI test: sqcli.exe refuses to start if the GUI is already open). For automation purposes, it would be relevant to know: can an MCP client work against the same running instance while the user is simultaneously active in the GUI — or do the two block each other? 2) Missing Functionality in the Current MCP Toolset (as of "first release") Currently available: list_projects, list_strategies, list_databanks, get_strategy_stats, run_project, stop_project. For our use case (automated snippet creation + trade evaluation), the following are missing: Snippet management via MCP: A tool to create/update Java snippets (currently only possible via direct filesystem access to the extend/Snippets/SQ/... folder + manual restart to trigger recompilation — no MCP tool exists for this) Detailed trade export via MCP: get_strategy_stats only returns aggregated metrics (Fitness, Net Profit, Number of Trades, Sharpe Ratio, Drawdown) — no individual trades. An MCP equivalent to the existing CLI commands -tools action=orderstocsv and -databank action=export would be needed to retrieve full trade lists programmatically Snippet reload without restart: Currently, a restart of the SQX instance appears necessary for new/changed Java snippets to be detected and recompiled (hash comparison on boot). An MCP tool for targeted reloading/recompiling of individual snippets without a full restart would be significantly more practical for automated workflows Robustness tests via MCP: Not currently mentioned in the toolset — Monte Carlo, System Parameter Permutation, and Walk-Forward runs can indirectly be triggered via the existing CLI through -project tasks, but a dedicated MCP tool with a structured return value (instead of plain log text) would be desirable 3) Open Question Regarding the Roadmap Is there a timeline for when the functionality listed above (snippet management, detailed trade export) will be added to the official MCP toolset

Dieter
StrategyQuant

MCP server (145 RC1): questions and requests for use with custom projects

The MCP server in 145 RC1 covers most of what I'd want an AI client to do: read configs, run projects, manage databanks. Before trusting it with a real mining pipeline, I'd value answers on these: Custom databank columns. Does query_databank, or a databank export through manage_databank, return custom columns from user column snippets and Custom Analysis, with the values the databank grid shows? Please check columns that override getValue() too: in my testing the native filter engine reads those as 0. (+1 to the existing request for custom columns and a bulk stats call.) Custom projects. Does run_project run every task of a multi-task custom project in sequence? Can get_project_config and set_task_config reach every task type (Filter with a Custom Analysis snippet, Monte Carlo cross-checks, SPP, Retester), or only Builder sections? Instrument definitions. Can a client read an instrument's full definition over MCP, including commission, swap and sessions? That's where cost errors hide. Bulk trades. Could get_strategy_trades have a bulk version for a whole databank? Correlation work on a 1,000-strategy pool is impractical one strategy at a time. Safety. Please consider a read-only mode and per-tool switches in SQX itself, so execute_shell, run_python, write_file and the delete tools can be turned off at the server. Is authentication planned for the endpoint? Can strategies deleted by filter_by_correlation or manage_databank be recovered? Concurrency. Is it safe to query databanks and configs while a long custom project is running? Engine parity. Do backtests started over MCP use the same engine path as the GUI? Does 145 change backtest results compared with 144 for the same strategy, data and settings? Stability. Will tool names and parameters stay the same from RC1 to final, and is there a rough date for 145 final? Thanks for building this.

Michael
StrategyQuant

move/copy databank strategies into another projects databank

Add feature to automatically move/copy databank strategies into another projects databank (finals collector), I am doing this now manually

mne1000
StrategyQuant

python export

I think strategyquantx should include python export. A realtime testing utilmwhere i can test the strategy in realtime from price feeds from the compatible platform. real time testing. the ability to backtest or strategy test strategy from strategy quant x inside the platforms without manualy testing them. A gui creator for my mql5 buisness strategys. The ability to add realtime swal data a tool for gathering information

alexander towell
StrategyQuant

Grid Martingale option in Builder Mode

Hello! It would be a great upgrade to add a grid martingale option available in builder mode, it will be much more siplier for us to add it from the option that tha algo wizard has.

A.R.S
StrategyQuant

How to Solve Look-forward behaviour in SQX

I realize that whenever SQX codes (with multi timeframe) pasted in MultiCharts or Tradestation, it instantly underperform badly. SQX uses look-forward data for higher timeframes when [0] is used. eg. Close[0], HighD[0]. This is highly disturbing and wasting resources especially GA to continue optimizing codes in that direction. I try to look into “What to Build“ > “# of Conditions, Period“ > “Global Lookback period Shift” There is no “Minimum“ for Shift. This look-forward bias generation problem could just be solved if SQX allows Higher timeframe data to specify its shift to Minimum = 1. This is a high ROI, low effort step SQX could do in next Build. Thanks.

marc
StrategyQuant

Option to hide buttons of "Long enabled" and "Short enabled" on business module

Add option to hide buttons of "Long enabled" and "Short enabled" on business module. For example, If you build a long only strategy, you don't want these buttons on the EA.

Frank
StrategyQuant

add Margin% field for calculations useful for CFDs and Futures

Fiz
StrategyQuant

Percent-of-price block for Stop/Limit Price Ranges

The “Stop/Limit Price Ranges” category offers only volatility-based blocks (ATR, BarRange, TrueRange etc.). Request: add a percent-of-price block to this category, usable by the Builder, so that a stop or limit entry can be placed x % before or after a level (e.g. 0.05 % below the daily high).

Daniel Kiefer

Next up

Committed and queued

No Next up issues

In Progress

Actively being built

No In Progress issues

Done

2

Recently shipped