Release · November 2026

BUILD ALPHA Update:
MCP + Prop Firm Simulator

Control Build Alpha with your AI agent. Connect Claude Desktop, Claude Code or Codex, then drive the interface with a prompt or run it headless. Plus a Prop Firm Simulator that ranks portfolios by the cash they are expected to pay you, and one-click data downloads from MetaTrader 5.

By David Bergstrom·Nov 11, 2026·11 min read

Watch the Demo See Licensing

Two big features in this release. Build Alpha MCP lets you control Build Alpha with your AI agent: connect Claude Desktop, Claude Code or Codex, then set up the interface with a prompt or run Build Alpha headless and never open it. The Prop Firm Simulator replays your portfolios through prop firm challenge and funded-account rules and ranks them by expected payout per attempt. You also get a built-in database of futures prop challenges with a new Find Challenges tool and a one-click MetaTrader 5 data downloader, plus fixes to the Python code for Interactive Brokers, the MT5 code for Darwinex and other brokers, and alternative data comparisons in NT8 and MT5 code.

Build Alpha MCP: Control Build Alpha With Your AI Agent

Build Alpha now has an MCP server. MCP (Model Context Protocol) is the open standard AI apps use to connect to outside tools. Connect Claude Desktop, Claude Code or Codex to Build Alpha and your agent can do nearly everything you do with a mouse: configure a search, build signals, run robustness tests and export code.

Just ask. A 40-second look at Build Alpha MCP in action.

There are two ways to use it.

Prompt the GUI

Set Up the Interface With a Prompt

Describe the search you want and the settings fill in inside Build Alpha. Forget where a setting is? Ask Claude about it. Then go back to the UI and continue your setup or work.

Headless

Run Build Alpha Without the Interface

Want agents to use Build Alpha and never look at the user interface? Go ahead. Run it headless and let the agent configure, run, filter and export on its own.

MCP vs. the LLM Orchestrator

In v3.1, the LLM Orchestrator let Build Alpha call an AI model to tune AutoTune between rounds. Build Alpha stayed in charge and the model advised. MCP flips that around: your agent is in charge, and Build Alpha is the research engine it calls. Use whichever fits how you work.

Getting Connected

The Build Alpha MCP server runs locally alongside Build Alpha, so use a desktop or command-line client: Claude Desktop, Claude Code or Codex (the Codex CLI, IDE extension and ChatGPT desktop app share one MCP configuration). Add the Build Alpha server to your client's MCP settings and start prompting.

What Your Agent Can Do in Build Alpha

Through MCP, an agent has access to the same research workflow you do.

MCP Capabilities

Configure a search end to end
Create custom signals or entries
Work with the full signal library
Add custom data and symbols
Run every search type: genetic search, portfolio search, prop firm simulation and portfolio simulation
Set Automated Filters and Workflows
Set Sequence Workflows so Build Alpha only surfaces strategies that pass your robustness gates
Export trade lists and strategy code

The last few matter most. Pair an agent with Sequence Workflows and your robustness gates, and Build Alpha only surfaces strategies that survive every test you set: out-of-sample, Vs Random, the Noise Test, Monte Carlo Permutation and more. The agent does the clicking. The gates do the judging.

Sample Prompts

Two examples from opposite ends of the research spectrum: a quick idea check and a large search with robustness gates.

Single Strategy: Test a Quick Idea

Prompt · Quick idea

Test this idea on ES 60-minute bars from 2010 to today: go long when the close is below the lowest close of the prior 5 bars and the 200-bar SMA is rising. Exit after 4 bars or at a 2 ATR stop. Show me the equity curve, profit factor, trade count and max drawdown, then run the Noise Test and Vs Random on it.

Run a Massive Search With Robustness Gates

Prompt · Overnight search

Run a genetic search on ES, NQ, CL and GC 60-minute bars from 2010 to one year ago from today. Use the full signal library plus COT and VIX contango data, max 3 rules, long strategies, fitness Net PNL. Set up a 3-round Sequence Workflow: round 1 keeps strategies with 300+ trades and positive P&LDD above 4; round 2 requires passing Vs Random and the Noise Test according to the Taguchi method blog; round 3 requires a final out-of-sample period on the last one year of data with passing strategies showing net profit. Advance when 500 strategies pass round 1 or after 8 hours. Export trade lists and NinjaTrader code for every survivor to my Research folder.

Forget Where a Setting Is? Ask.

Prompt · Quick question

Where do I change the date range and out-of-sample period for round 2 of my Sequence Workflow?

Then go back to the UI and keep working. Or never touch the UI at all.

Trading research on a new level.

Prop Firm Simulator: Rank Portfolios by Payout, Not Pass Rate

The requests for prop firm features from Build Alpha users became too overwhelming to ignore. Many of you are already killing it with these firms. The Prop Firm Simulator replays your portfolios, bar by bar, through a prop firm's challenge and funded-account rules from every historical start date, then reports what a prop trader actually gets paid on.

Here is why that matters. In a simple example from the full guide, a strategy passes the evaluation 59% of the time but collects a payout only 35% of the time. Passing is not getting paid. So the simulator ranks portfolios by Expected Net / Attempt: what one challenge purchase is worth after every fee.

A 42-second walkthrough of the Prop Firm Simulator.

Build Alpha Prop Firm Simulator settings: choosing a futures prop firm and challenge from the built-in database and loading its rules — Pick a prop firm and challenge from the built-in database, click Load, and its rules fill in.

Challenge, Funded and Simulation tabs. Profit targets, trailing and end-of-day drawdowns, daily loss limits, contract limits, minimum trading days, monthly fees, resets, consistency caps and payout rules.
Portfolio generation. Searches combinations of your saved strategies with min and max strategy counts, a minimum number of symbols and a correlation cap.
Fixed contracts, no compounding. Micros count 0.1 toward the contract limit, and an entry that would exceed it is skipped, never trimmed.
Rule set files. Import and Export one file that holds the challenge, funded and simulation settings together.

New Prop Fitness Functions

Expected Net / Attempt (recommended): what one challenge purchase is worth after fees
Prop ROI: trader payouts per attempt divided by cost per attempt
Pass Probability: share of start dates that hit the target before breaching
Survival Prop: favors funded accounts still open at the end of the horizon

Build Alpha Prop Firm Simulator results window ranking portfolios by pass probability, payout metrics, Expected Net per Attempt and Prop ROI — Challenge, Funded and Fitness metrics side by side for every candidate portfolio.

In those results, the portfolio with the highest pass rate (90.7%) is worth +$1,785 per attempt. A portfolio further down passes only 73.2% of the time but is worth +$2,659 per attempt with a 7.53x Prop ROI, because it reaches payouts sooner and more often.

Build Alpha Prop Firm Simulator results sorted by pass probability, payout probability and Expected Net per attempt, then Account Paths with hover on each path — Sort by Pass Prob, then 1st Payout Prob, then Expected Net / Attempt. Open Account Paths and hover any run.

Click any portfolio, open Account Paths and you get one line per funded start date. Hover a line to see that run's payouts, gross, net to the trader and outcome. And when you generate code, the prop logic carries into it: maximum counts, position stacking, group stacking across custom symbol groups, and sizing.

Cornerstone Feature

Optimize for the Payout, Not the Pass.

The full guide walks through the prop firm math (why your cash flow is not your strategy's P&L), the expected value formula, real payout rules like qualifying days and caps, every metric in the results window, and why the simulator replays real start dates instead of shuffling trades.

Read the full Prop Firm Simulator guide

Find the Best Challenge for Your Portfolio

Preconfigured prop firm rules are already here. Build Alpha now includes a database of futures prop firm challenges, and an AI agent scans the web to keep it current with the latest challenges, rules and pricing. As of this writing it holds 800+ active challenges from 42 firms. Pick a firm and a challenge in the simulator settings, click Load, and its rules fill in. No typing required.

Then work the other direction with Find Challenges, the prop firm Challenge Finder, available from the right-hand panel of the Prop Firm Simulator results. Select a portfolio and it runs that portfolio through every challenge in the database, ranked in a sortable table by Expected Net / Attempt. View any challenge's Account Paths for that portfolio, or load that challenge back into the Prop Firm Simulator to search for better portfolios built for it. Firm rules the simulator can't model, like automation, overnight and news rules, get their own columns so you can check them before you buy.

Build Alpha Find Challenges ranking every futures prop firm challenge for one portfolio by Expected Net per Attempt, then opening Account Paths — One portfolio ranked across every challenge in the database, then Account Paths for two of them.

Run This Before You Commit

The same portfolio may work for one challenge and be a disaster for another. Run Find Challenges before you buy a challenge to make sure you're optimizing your spend and your chances at a payout.

MetaTrader 5 Data Downloader

Build Alpha can now pull data straight from MetaTrader 5. Go to File → Tools → Settings → Add from MT5, search for a symbol, choose your timeframes and timezone, and click Download.

Build Alpha MT5 Data Downloader window with symbol search, timeframe and timezone selection — Search any MT5 symbol, pick timeframes and a timezone, and download it straight into Build Alpha.

Setup

One-Click Connector

Install the Build Alpha connector into MetaTrader 5 with one click. No scripts to copy and no files to move.

Contract Specs

Every Detail Comes Across

Tick size, point value, margin and swaps download with the price data, so backtests use your broker's real contract settings.

Advanced Backtesting

Slippage and Tick Data

Optionally download slippage and tick data for Build Alpha's advanced backtesting modes.

Downloaded symbols are formatted for Build Alpha and added to your Symbols dropdown instantly, ready to test. Because the data comes from your own MT5 broker, you build and test on the same data you trade. See Build Alpha for MetaTrader for the full MT5 workflow.

Code Export Fixes: IBKR Python, MT5 and NT8

Three fixes for traders running Build Alpha strategies live.

Python · Interactive Brokers

Fixed Python Code for IBKR

The Python code Build Alpha generates for Interactive Brokers has been fixed so live signals line up with the backtest, including indexing and timezone handling in the Python library.

MetaTrader 5 · Darwinex

Fixed MT5 Code for Darwinex and Other Brokers

Generated MT5 code now runs correctly on Darwinex and other MetaTrader 5 brokers, with a bug-fix pass on the MT5 strategy library and bar timestamps aligned to Build Alpha's backtest.

NinjaTrader 8 · MT5

No More Rounding Issues on Alternative Data

Comparisons against alternative data in generated NinjaTrader 8 and MT5 code no longer hit rounding issues, so a condition like a COT index above 0.6 evaluates the same way it does in Build Alpha's backtest.

See the Python trading strategy builder, the MetaTrader EA generator and Build Alpha for NinjaTrader for more on each platform.

Coming Next

A Live Execution Layer

Streaming P&L, live position tracking, and native execution coming directly to Build Alpha in the next release. The gap between strategy generation and live trading gets one step shorter.

Build Alpha live execution preview with streaming P&L and real-time position tracking

How to Get the Update

Check your email! It is live if you're reading this. If you didn't get it please check spam then email me.

If you're new to Build Alpha and reading this from outside the user base, the demo is the fastest way to see what Build Alpha can do. To see licensing please head to the Licensing Page.

Build Alpha on a desktop monitor, connected to OpenAI, Claude, Gemini and Grok

Build Alpha: generate, validate and export strategies, now with your AI agent at the controls

David Bergstrom

Founder, Build Alpha

A decade-plus in professional trading as a market maker and quantitative strategy developer at a high-frequency trading firm with a CME seat. Consulting for hedge funds, CTAs, family offices, and RIAs. Self-taught C++, C#, and Python programmer specializing in data science, machine learning, and trading strategy development.

Frequently Asked Questions

What's new in this Build Alpha update?

Build Alpha MCP lets you control Build Alpha with your AI agent through Claude Desktop, Claude Code or Codex. The Prop Firm Simulator ranks portfolios by expected payout per attempt under prop firm rules, with a built-in database of futures prop challenges and a Find Challenges tool. A new MetaTrader 5 data downloader pulls symbols, contract specs and optional tick data straight from your MT5 broker. The Python code for Interactive Brokers, the MT5 code for Darwinex and other brokers, and alternative data comparisons in NT8 and MT5 code were also fixed. A new MetaTrader 5 data downloader pulls symbols, contract specs, and optional slippage and tick data straight from MT5.

Can Build Alpha download data from MetaTrader 5?

Yes. Go to File → Tools → Settings → Add from MT5 and install the connector with one click. Then search for a symbol, choose timeframes and a timezone, and click Download. Build Alpha pulls the price data plus tick size, point value, margin and swaps, can also download slippage and tick data, and adds the symbol to your Symbols dropdown.

What is the Build Alpha MCP?

An MCP server that connects AI agents to Build Alpha. MCP (Model Context Protocol) is the open standard AI apps use to connect to outside tools. Once connected, an agent can configure searches, build signals, run robustness tests, set workflows and export code inside Build Alpha.

Which AI apps can connect to Build Alpha?

Claude Desktop, Claude Code and Codex (the Codex CLI, IDE extension and ChatGPT desktop app share one MCP configuration). The server runs locally alongside Build Alpha, so use a desktop or command-line client. Other clients that support local MCP servers can connect the same way.

Can I run Build Alpha headless with an AI agent?

Yes. You can let the agent set up the interface while you watch, or run Build Alpha headless so the agent configures, runs, filters and exports without the interface ever opening.

What can an agent do in Build Alpha through MCP?

Configure a search end to end, create custom signals or entries, work with the full signal library, add custom data and symbols, run genetic search, portfolio search, prop firm simulation and portfolio simulation, set Automated Filters, Workflows and Sequence Workflows, and export trade lists and strategy code.

How is MCP different from the LLM Orchestrator?

With the LLM Orchestrator, Build Alpha calls an AI model to tune AutoTune between rounds, so Build Alpha stays in charge. With MCP, your agent is in charge and calls Build Alpha as its research engine. Use whichever fits how you work.

I forgot where a setting is. Can the agent help?

Yes. Ask Claude where the setting lives or what it does, then go back to the interface and continue your setup or work.

Does the AI trade live?

No. The agent works during research. Strategies export as native code to platforms like NinjaTrader, TradeStation, MultiCharts, MetaTrader, TradingView and Python for Interactive Brokers, and run as deterministic code with no dependency on an AI connection.

What is the Prop Firm Simulator?

A tool that replays portfolios bar by bar through a prop firm's challenge and funded-account rules from every historical start date. It reports pass probability, payout probability, payouts, survival and the expected net value of one challenge purchase after all fees. See the full Prop Firm Simulator guide.

What is Find Challenges?

A tool in the Prop Firm Simulator results that runs a selected portfolio through every futures prop challenge in Build Alpha's database and ranks them in a sortable table by Expected Net / Attempt. An AI agent keeps the database current with the latest challenges, rules and pricing. You can view any challenge's Account Paths or load that challenge back into the simulator.

Is live trading coming to Build Alpha?

Yes. A live execution layer with streaming P&L, live position tracking and native execution is coming in the next release.

Which fitness function should I use for prop firm portfolios?

Expected Net / Attempt is the recommended choice because it measures what one challenge purchase is worth after every fee. The highest pass rate is often not the portfolio that pays the most.

What code export fixes are in this update?

The Python code generated for Interactive Brokers was fixed so live signals line up with the backtest, the MT5 code now runs correctly on Darwinex and other MetaTrader 5 brokers, and alternative data comparisons in NT8 and MT5 code no longer have rounding issues.

Is there an extra cost for these features?

No. MCP and the Prop Firm Simulator are included with your Build Alpha license. The only outside cost is your own Claude or ChatGPT plan or API usage for the agent you connect.

How do I update Build Alpha to the latest version?

Check your email! New members head to the Licensing Page.

Does Build Alpha run on Mac or Linux?

Build Alpha is a native Windows application. Mac and Linux users run it under Parallels, VMware or Boot Camp. Windows 10 and Windows 11 are supported.