KnitPkg brings dependency management to MQL4/MQL5 by replacing folder copying and manual #include edits with Git-backed, SemVer-versioned components declared in a manifest. A lock file pins exact revisions, making installs and builds repeatable across terminals and machines.

Projects define dependencies in knitpkg.yaml (YAML/JSON). KnitPkg installs them as either a classic include tree or a single generated header for flat distribution, while a registry stores metadata and resolves version ranges to specific Git commits.

Composite packages handle external dependencies cleanly using generated autocomplete headers for development and install-time directives that KnitPkg converts into real includes. The workflow is validated with unit tests compiled from the manifest, demonstrated by reusing one SMA implementation across a package, indicator, and CPU-effici...

👉 Read | NeuroBook | @mql5dev

#MQL5 #MT5 #AlgoTrading

A multi-timeframe trend indicator consolidates MA-based direction from M1, M5, M15, and H1 into a single trend state: bullish, bearish, or sideways. The aggregation is designed to reduce conflicting signals by weighting aligned timeframes and flagging mixed conditions as neutral.

Sideways detection is handled with dot markers that appear when MA behavior suggests low directional strength, helping filter entries during range conditions. Additional arrow markers provide entry timing cues on a selected timeframe such as M30, indicating when conditions align for a potential position.

A separate MA-based trend readout for the active chart timeframe is included to keep the current context visible alongside the combined MTF signal.

👉 Read | Quotes | @mql5dev

#MQL5 #MT5 #Indicator

Cluster-based classification can be applied to indicator slope to separate flat markets from trending phases using numeric thresholds rather than visual judgment.

An MQL-style HalfTrend example evaluates only x[iB] - x[iB+1], normalized to _Point for cross-symbol comparability. With a two-buffer line (up/down), the active buffer value becomes the source series, and fabs(delta) avoids duplicating clusters by sign.

Instead of k-means with prohibitive cost at scale, the approach uses incremental weighted mean and variance (Finch) in a single pass, enabling fast categorization over full history. Clustering is isolated in ClusterTrend.mqh, controlled by NumValues, and benchmarked in milliseconds on multi-thousand bar datasets.

👉 Read | Quotes | @mql5dev

#MQL5 #MT5 #AI

CExecutionSafety is a small MQL5 include class intended to stop Expert Advisors from placing orders when execution conditions degrade. It tracks two latency components in real time: terminal ping and measured execution delay around trading calls, then blocks trading when the combined value exceeds a configured limit.

Ping is read from TERMINAL_PING_LAST, while RecordExecDelay() stores a wall-clock delta taken around a CTrade operation. CheckExecutionSafety() sums both values, compares against the threshold, and logs diagnostics when the gate fails.

If the terminal is disconnected, TERMINAL_PING_LAST returns -1. This is treated as an automatic failure to prevent orders without server contact.

Read-only accessors expose ping, last execution delay, combined latency (9999 when disconnected), sample availability, and the configured threshold. Changelog: v1.0 init...

👉 Read | Docs | @mql5dev

#MQL5 #MT5 #EA

CLatencyMonitor is a lightweight MQL5 include-file that adds latency awareness to execution-sensitive EAs via a single reusable class. It measures inter-tick deltas with GetTickCount64(), applies an ATR-based volatility gate to reduce false positives during high-volatility bursts, and publishes confirmed lag events through a named GlobalVariable for cross-EA IPC on the same terminal.

On each OnTick() call, the elapsed milliseconds since the previous tick are computed, with the first tick seeding state only. A lag candidate is suppressed when ATR exceeds AtrSmaMultiplier × SMA(ATR, AtrPeriod), with the SMA calculated manually from the ATR buffer for self-normalisation across symbols and timeframes. If ATR setup fails or history is insufficient, the gate fails open so detection remains active.

A timer-driven persistence check confirms lag only after Persistenc...

👉 Read | Quotes | @mql5dev

#MQL5 #MT5 #EA

ExMachina Safe Scalping is a breakout scalping EA designed to stay flat unless six independent conditions align on the same bar. Filters include dual EMA trend (150/510), ATR-based trend strength to reject weak ranges, price location relative to both EMAs, an N-bar high/low breakout with a volatility-adjusted buffer, RSI kept out of extremes, and close-to-close momentum confirmation.

Risk controls are fixed per trade: predefined stop loss and take profit at entry, single open position, and no martingale, grid, hedging, or averaging down. Breakeven can move the stop to entry after favorable movement, and a drawdown threshold can auto-pause trading.

Operational filters include session-hour limits, maximum spread gating, and optional news-time blocking. The Code Base release is a single .mq5 file using the standard MT5 Trade library, with readable logic and an ...

👉 Read | Forum | @mql5dev

#MQL5 #MT5 #Scalper

The DoEasy chart subsystem is finished by making chart windows first-class library objects, enabling reliable auto-refresh of chart and subwindow properties. Instead of manually syncing state, every tracked integer/real/string property updates its stored parameters and emits structured custom events when values change, cross a threshold, or match a configured combination.

Core refactor work includes new message indices, a dedicated chart-window list ID for event attribution, and aligned enumerations for properties and sort criteria so search/sort stays consistent. The chart window class now owns its property arrays, descriptions, compare logic, and refresh flow, including indicator add/remove handling.

Chart objects gain symbol/timeframe change events, explicit helpers to set tracked values, and cleaner property population methods. Practical result: EA...

👉 Read | Signals | @mql5dev

#MQL5 #MT5 #AlgoTrading

Last Structure Indicator (Single Phase) implements a framework to test the Last Structure Break (LSB) concept using structured price-action rules.

Session boundaries are detected via logical day separation, based on chart separators and datetime normalization rather than daily timeframe data. This keeps day transitions consistent across timeframes and broker feeds.

The scan builds support/resistance pivots with configurable modes (wick-based or body-based) and pivot strength validation using left/right bar requirements. Only the final structural levels of a session that remain unbroken into the close are accepted, then projected forward as trendlines into the next session.

Designed for real-time use with minimal recalculation, the tool aims to measure whether end-of-session structure has repeatable relevance for subsequent price action.

👉 Read | Forum | @mql5dev

#MQL5 #MT5 #Indicator

This article breaks down how to consistently identify Premium/Discount (PD) arrays and imbalances in price action, and why traders fail when they pick the wrong “reaction zones” or trade against the broader narrative. It emphasizes using a simple checklist mindset to reduce impulse entries and indecision.

Core concepts are defined practically: imbalances as fast, one-sided moves that often get revisited; market structure shifts as decisive breaks that flip highs/lows; PD arrays as splitting an expansion leg into premium vs discount to avoid buying high or selling low; and using fractal behavior to refine higher-timeframe ideas into lower-timeframe entries.

A MetaTrader 5 EA in MQL5 is presented to automate this workflow: detect trend and consolidation with stacked moving averages, map higher-timeframe expansion legs into PD zones, track imbalances inside th...

👉 Read | VPS | @mql5dev

#MQL5 #MT5 #EA

An updated Limit Channels indicator adds a “states” panel rendered as a colored histogram to summarize channel conditions.

States are derived from three nested level groups. Level 3 uses the outer bands and represents the slowest state. Level 2 uses the middle bands. Level 1 uses the inner bands and reacts fastest.

The six total levels allow additional combinations beyond the default three-state mapping. Color transitions in the histogram can be used as event markers, with signal rules left to the implementer based on timeframe, instrument, and risk constraints.

👉 Read | Freelance | @mql5dev

#MQL5 #MT5 #Indicator

ALGLIB’s MQL5 port ships minLBFGS but lacks a box-constrained variant, pushing developers toward heavier optimizers for simple bound-limited calibration tasks. A compact Truncated Newton Conjugate-Gradient (TNC) solver fills that gap with an API designed to drop into projects like minLBFGS.

TNC approximates Newton steps using an inner Conjugate Gradient loop, relying on Hessian-vector products derived from gradient finite differences instead of forming a full Hessian. It supports analytic gradients when available, plus stable numerical differentiation near bounds, while enforcing constraints via projection and variable clamping.

The implementation is packaged as tnc.mqh with a CFunctor/CTruncNewtonCG interface, tunable stopping criteria, and clear termination codes. It’s validated on Rosenbrock and used to fit logistic regression alongside LBFGS, pro...

👉 Read | AlgoBook | @mql5dev

#MQL5 #MT5 #algorithm

A trend indicator is used to classify market direction as uptrend or downtrend based on the line state.

Common operational logic is straightforward: a green line is treated as a bullish condition and a red line as a bearish condition. Some traders use this as a basic ruleset to align entries and exits with the prevailing direction.

For practical use, results depend on the instrument, timeframe, and parameter settings. As with any single-signal approach, confirmation with volatility, volume, or higher-timeframe structure is typically required to reduce false signals.

👉 Read | Freelance | @mql5dev

#MQL5 #MT5 #Indicator

This MQL5 binomial distribution viewer evolves from a 2D canvas into a Direct3D-backed tool that can switch between 2D and 3D at runtime. The goal is practical: make frequency differences and PMF shape easier to inspect using depth, perspective, and camera control instead of restarting or redrawing separate views.

The implementation adds CCanvas3D plus DXBox primitives, a view-mode enum, and configurable inputs for camera, ground plane, and axis styling. The codebase is refactored into a single visualizer class that centralizes state, data arrays, and interaction logic, with a destructor that explicitly releases GPU resources.

On the 3D side, histogram bins become box meshes updated via scaling/translation matrices, with an auto-fit camera derived from scene bounds. Mouse interactions handle rotation and zoom, while 2D overlays (header, legend, swit...

👉 Read | AppStore | @mql5dev

#MQL5 #MT5 #AlgoTrading

An updated version of the MetaTrader 4 platform will be released on Friday, March 13, 2026. This build includes improved interface translations, bug fixes, and enhanced platform stability.

The update will be distributed via Live Update.

Discuss the update...

An MT4 version of the Limit Channels indicator has been published.

The release targets traders and developers maintaining MT4 stacks that require channel-based boundary calculations with configurable limits. Typical usage includes defining dynamic upper and lower bands and tracking price interaction with those levels for alerts or signal logic.

Implementation details, parameters, and example configurations are available at the referenced source.

👉 Read | Quotes | @mql5dev

#MQL4 #MT4 #Indicator

Symbol discipline fails when a solid strategy is applied to the wrong instrument. This MQL5 solution enforces a strict symbol whitelist at the execution layer, blocking both manual and EA trades on non-approved symbols using OnTradeTransaction.

The design is modular: a shared include file centralizes whitelist parsing, case-insensitive validation, and file I/O; an indicator dashboard writes the approved universe and displays status plus recent blocks; an enforcement EA reacts to ORDER_ADD and DEAL_ADD events to cancel pending orders or flatten positions.

Configuration is persisted in a plain text file, blocked actions are audited to CSV, and components synchronize via shared storage. Result: strategy scope stays aligned with backtests, exposure can’t silently expand, and every violation is traceable.

👉 Read | CodeBase | @mql5dev

#MQL5 #MT5 #AlgoTrading

A minimal channel tool can be built from recent high/low values with band spacing derived from ATR. The result is a dynamic range framework that adapts to volatility and updates with each bar.

In practice, these bands are more suitable for estimating short-term support and resistance than for generating entries. They can also serve as consistent reference levels for take-profit and stop-loss placement in systems that already define direction and timing.

This setup is intended for trend context, range boundaries, and risk parameterization, not as a standalone signal source.

👉 Read | NeuroBook | @mql5dev

#MQL5 #MT5 #Indicator

This article turns discretionary price action into a testable model by representing confirmed swing highs/lows as nodes and swing transitions as edges in a market-structure graph. Depth-First Search is then used to “commit” to one structural branch (higher highs/lows or lower lows/highs), only accepting a directional bias after a minimum path depth confirms continuation.

The EA implements this in MQL5 with configurable swing sensitivity, minimum depth, target tolerance, and either fixed-lot or risk-based sizing. Key invalidation levels (last higher low / last lower high) make trend failure deterministic: if broken, the system backtracks, clears state, and searches for an alternate path.

Operationally, the design emphasizes longevity: daily resets, capped node storage, array resizing, and cleanup on deinit keep the swing graph and DFS state stable d...

👉 Read | AlgoBook | @mql5dev

#MQL5 #MT5 #AlgoTrading

Quant research keeps producing backtests with high Sharpe, low drawdown, and immediate live failure. The main driver is overfitting through multiple channels, not a single mistake fixed by a basic train/test split.

Common failure modes include data snooping across many candidates, curve-fitting to path-specific noise, and accumulated researcher degrees of freedom during iterative tuning.

Three controls target different layers. Validation-within-Validation enforces outer training, inner validation, and a one-time final test, often with anchored walkforward. CPCV removes temporal leakage via purging and embargoing, then evaluates combinatorial OOS paths. CSCV audits selection with Probability of Backtest Overfitting, reporting how often the in-sample winner ranks poorly out-of-sample.

👉 Read | CodeBase | @mql5dev

#MQL5 #MT5 #AlgoTrading

The article shifts a retail stat-arb workflow from SQLite (OLTP) to DuckDB-backed analytics using Parquet files, keeping mean-reversion and sub-second HFT out of scope. The goal is faster research loops: qualify ideas in seconds before spending time on full MT5 EA backtests.

It shows how to export an entire SQLite database or selected tables to Parquet via DuckDB, then query those files directly with SQL or Python’s lazy relational API.

Key technical gains come from Parquet’s columnar layout, embedded min/max statistics, compression, and DuckDB’s zero-copy reads, which reduce I/O and skip irrelevant files.

For scalable history storage, the plan is to denormalize price data and adopt Hive-style partitioning (source, asset class, ticker, timeframe, year, month) to keep datasets portable and easy to merge across brokers and providers.

👉 Read | NeuroBook | @mql5dev

#MQL5 #MT5 #AlgoTrading

ExMachina Trade Pilot is an MT5 expert advisor focused on order execution and position lifecycle management from a single control panel. It targets common platform gaps: partial take-profits, automated breakeven, and trailing logic beyond fixed-point trails.

The workflow consolidates risk-based lot sizing, SL/TP placement, and multi-stage exits. Up to three take-profit levels can be configured with independent close percentages; each hit triggers an automatic partial close with logging and live status tracking.

Management tools include ATR-based, fixed-point, and previous-candle trailing with a minimum step, plus breakeven on a profit threshold with optional offset. One-click market and pending orders apply predefined offsets and SL/TP rules. Additional controls cover close-all, direction-based closes, pending cleanup, and a dashboard for spread, exposure, ...

👉 Read | Calendar | @mql5dev

#MQL5 #MT5 #EA