Meta

The following sections describe the latest updates to the documentation.

2026-04-28

• [Source] Introduced a new dual-language SKILL system and build pipeline: added bundle-skills.py (replacing validate-and-install-skills.py) to validate frontmatter, split unified SKILL.md sources into Python/C# variants at build time, bundle skills.zip, optionally install via --py/--cs, enforce duplicate-name checks, and reorganize sources into skill-templates/. Updated CI to run the new bundler.
• [Source] Overhauled Notifications: merged Live Commands and Custom Signal Exports into a unified external-communication router under Notifications, clarifying when to use Signal Exports vs notify.web and Commands vs Object Store. Converted the content to dual-language.
• [Source] Converted the Indicator Universes guidance to dual-language and reinforced best practices (per-symbol SelectionData lifecycle, daily selection cadence, Equity price_scale_factor resets, Crypto TradeBar wrapping).
• [Source] Converted the Logging skill to dual-language and tightened guidance on channels, quotas, rate limits, live-only safe_log, and using Object Store for structured diagnostics.
• [Source] Removed/retired several legacy SKILL pages from the '03 Writing Algorithms' tree as part of the move to skill-templates/, including the Charting SKILL.
• [Source] Deleted the Language Selection SKILL; language handling is now implicit in the new dual-language SKILL system.

2026-04-27

• [Source] Increased QCC pricing for QuantConnect Futures data downloads across all resolutions (Tick 150 QCC/$1.50, Second 100 QCC/$1.00, Minute 50 QCC/$0.50, Hour 1,500 QCC/$15, Daily 1,500 QCC/$15) and updated cost examples accordingly.
• [Source] Added a Notifications SKILL with best practices for Notify.* (cloud live-only, quotas/limits, channel-specific guidance, event-based triggering, and data-usage rules), later streamlined and cross-linked to use signal exports (not webhooks) for sending portfolio targets.
• [Source] Introduced a comprehensive Live Commands SKILL covering generic vs typed handlers, live-only behavior, sending via link/broadcast/REST/CLI, cross-project coordination, return semantics, and C# equivalents.
• [Source] Expanded the 'Send Commands by API' docs into step-by-step C# and Python guides with runnable scripts, CLI arguments, proper authentication headers, and payload casing.
• [Source] Expanded the 'Broadcast Commands by API' docs into step-by-step C# and Python guides with runnable scripts, CLI arguments, and organization-wide broadcast instructions.
• [Source] Added a Custom Signal Export SKILL for implementing ISignalExportTarget providers, emphasizing payload design for external endpoints, handling PortfolioTarget weights vs shares, batching, error handling, and HTTP client reuse.

2026-04-24

• [Source] Added Memory Metrics to Debugging Tools: document QuantConnect.OS metrics (ApplicationMemoryUsed, TotalPhysicalMemoryUsed, CpuUsage, GetServerStatistics), usage examples, and LEAN’s EMA-smoothed memory monitoring and termination behavior.
• [Source] Overhauled Scheduled Events guidance: new scheduling patterns, two constraints for rebalancing (seed_initial_prices and filtering zero-price securities), strict “don’t fire inside a forming bar” rule, timezone/ticker guidance, and common mistakes with C# mappings.
• [Source] Introduced Indicator Universes SKILL: end-to-end pattern for per-symbol indicators in universe selection (Equities and Crypto), including SelectionData, lifecycle cleanup, warm-up, Equity corporate-action resets with SCALED_RAW history, crypto OHLC handling, checklists, and C# implementation.
• [Source] Added Logging SKILL: proper use of log/debug/error/quit, per-tier log quotas, duplicate suppression on log, 1/sec + 200-char limits on debug/error, live-only safe_log, using order tags for context, Object Store logging for research, and best practices.
• [Source] Added IB FIX Integration docs: step-by-step IB Client Portal subscription to enable QuantConnect FIX routing and link your QuantConnect username.
• [Source] Added dedicated FIX live deployment page for Interactive Brokers, mirroring deploy steps with FIX-specific authentication details.
• [Source] Added Charting SKILL: series types and indices (overlay vs subcharts), reserved names to avoid, series and data-point quotas by tier, plot vs plot_indicator guidance by resolution, and candlestick plotting requirements.
• [Source] Added Language Selection SKILL: reliably detect and silently honor project language (Python vs C#) across skills; note the main.py entrypoint exception for Python projects.

2026-04-22

• [Source] Added a new Scheduled Events SKILL documentation for LEAN with best practices for scheduling rebalances: avoid firing during active bars, use 08:00 ET for dynamic US Equity universes on daily data, align universe selection with rebalance date rules, handle time zones (e.g., UTC for Crypto) and multi-asset scheduling intersections, allow string tickers in rules, and includes common pitfalls, a checklist, and C# equivalents.

2026-04-16

• [Source] Switched C# research docs to use NuGet-based Plotly.NET packages, added Plotly.NET.Interactive, replaced HTML(...toChartHTML(...)) with display(chart), and improved Jupyter cell separation across examples.
• [Source] Added explicit link to Equinix NY7 for co-located servers across security, live trading, deployment, and AI assistance pages.
• [Source] Fixed broken Parameters docs link in the Live Reconciliation OOS backtest guide.

2026-04-15

• [Source] Introduced a new Live Reconciliation section in Research Environment Meta Analysis, moving reconciliation content off Live Analysis and adding a complete OOS workflow: read live deployment parameters, run an out-of-sample backtest, overlay equity curves and order fills, and provide worked examples and metrics.
• [Source] Added C# code samples across the Live Reconciliation steps (reading live parameters, compiling/running OOS backtests, reading charts), and clarified DTW scoring is Python-only.
• [Source] Fixed order retrieval in Live Reconciliation tutorial by paginating ReadLiveOrders/ReadBacktestOrders in 100-ID windows and retrying the first window, preventing truncation beyond 100 orders.
• [Source] Expanded MCP Server Tools documentation with new Research (Jupyter cell/notebook ops), News (blog posts and web get), Environment (library support, datasets), and Notifications (email/SMS/Telegram) tools.
• [Source] Renumbered Live Analysis pages and shifted Live Deployment Automation to section 06; updated inbound link fragments (e.g., Plot Order Fills is now #03-Plot-Order-Fills).

2026-04-09

• [Source] Added a new Debugging > Order Tagging page with C#/Python examples showing how to tag orders for diagnostics and where tags appear in result files; linked from the order properties Tags section.
• [Source] Updated Lean CLI dataset download docs: removed CoinAPI datasets from the CLI flow (cannot be downloaded via CLI), deleted Crypto and Crypto Future cost pages and renumbered others; revised examples and added the historical data provider selection step.
• [Source] Fixed OpenAPI schema issues in Project endpoints: corrected types (e.g., maxFileSize to integer), and explicitly defined Grid, LiveForm, EncryptionKey, and BrokerageData models with clearer descriptions.
• [Source] Backtest API docs cleanup: simplified rollingWindow to a generic object, set totalPerformance to AlgorithmPerformance, and pruned overly verbose nested examples while consolidating model docs.
• [Source] Optimization Read endpoint enhanced with new metadata fields: created (timestamp), psr, sharpeRatio, trades, and cloneId; examples updated accordingly.

2026-04-08

• [Source] Added a new joblib-based workflow for transporting binary models using Base64 encoding and clarified pickle-based steps.
• [Source] Updated Bulk Download example to load a Base64-encoded joblib model from external storage and save a Base64 version to the Object Store; minor code cleanups.
• [Source] Backtest Create API response now includes an 'analysis' array and documented AnalysisResult model with sample and solutions.
• [Source] Backtest Read (Statistics) API response now includes an 'analysis' array and documented AnalysisResult model with sample and solutions.
• [Source] Equity Options example increased initial cash to 1,000,000 and added a quantity guard before placing the short straddle; copy updated with relevant links.
• [Source] Risk Management example adds required sklearn and xgboost imports to the XGBoost-based Alpha/Bracket risk model sample.

2026-04-07

• [Source] Morningstar US Fundamentals: added new classification enums (StockType, StyleBox, MorningstarEconomySphereCode) with examples, updated filtering guidance to use enumeration helpers, and corrected IndustryGroupCode example to ApplicationSoftware.
• [Source] Quiver WallStreetBets: noted the dataset ends in February 2025 and clarified it’s free to use on QuantConnect Cloud for research.
• [Source] Quiver WallStreetBets: added an End Date (February 2025) row to the data summary table and improved formatting.
• [Source] API Authentication: overhauled examples to use environment variables and a header helper (C# and Python), switched C# to HttpClient, and added authenticated cURL and PowerShell request examples.
• [Source] Equity Options Handling Data: added Example 4 (Delta-Hedged Short Straddle) with C#/Python implementations, including hedging helper classes; later refined dates/cash and formatting.
• [Source] Custom Indicators: added Example 3 (Custom Annualized Volatility Indicator) with universe shorting logic and CRSI-based exits; followed by formatting cleanup and consolidated C# field declarations.
• [Source] Risk Management Framework: added Python-only Example 3 (Bracket Risk Model) combining trailing stop-loss and take-profit; wrapped content in a Python-only section.
• [Source] Indicators (MOSC): corrected McClellan Oscillator examples to use default periods (removed explicit 19/39) in usage and history sections.
• [Source] Indicators (NHNL/NHNLV): fixed examples to include the period parameter (2) across usage, universe-based, and history sections.

2026-04-06

• [Source] Added new 'Example for Logging' section to Object Store docs (Writing Algorithms) showing how to log order data to a file and analyze it in Research.
• [Source] Added new 'Example for Logging' section to Object Store docs (Research Environment).
• [Source] Updated Python examples to use BrokerageName.QUANT_CONNECT_BROKERAGE in the QuantConnect Paper Trading demo algorithm.
• [Source] Updated Python examples to use BrokerageName.QUANT_CONNECT_BROKERAGE in the QuantConnect Paper Trading Supported Models intro.
• [Source] Corrected options pricing model example to inherit from CurrentPriceOptionPriceModel.

2026-04-02

• [Source] Added new annotated 1200x630 payoff charts and trade-setup explanations across all 41 option strategy pages; updated CDN image URLs.
• [Source] Fixed currency rendering by wrapping dollar amounts to prevent LaTeX/math conflicts across the updated strategy pages.
• [Source] Corrected Short Call Butterfly example P/L from a -$1,252 loss to a +$1,244 gain.
• [Source] Updated Short Straddle example to show a -$5,803 loss (was -$5,277).