Repertoire Builder 1.4 (Jan 2026 Release)

What started as a simple idea to help players build, understand, and train their opening repertoires has grown into a much deeper and more connected study platform. Over the past months, I’ve been steadily working toward the one year anniversary in April, refining core ideas, improving training workflows, and adding features that genuinely help with real chess improvement.

TLDR
You can use it for free here:
https://chessboardmagic.com/repertoirebuilder

New Features

Version 1.4 is a big step forward. It introduces new ways to build repertoires from your own games, apply modern AI to improve existing lines, train more realistically, and understand positions more clearly. Below is a walkthrough of what’s new in this release.

Repertoire Generator Build Openings From Your Own Games

A brand new Repertoire Generator has been added, allowing you to generate opening repertoires directly from your own game history.
You can import games from Lichess, Chess.com, or PGN files, and Chessboard Magic will automatically extract and structure the openings you actually play. Instead of starting from abstract theory or copying someone else’s repertoire, you can now build repertoires based on your real decisions and the positions you repeatedly reach in your own games.
Each generated repertoire can be previewed before being added, so you stay fully in control of what enters your study library. The Repertoire Generator is available both as a standalone application and fully integrated into the Repertoire Builder, making it easy to turn raw game data into structured, study ready openings.

Standard Puzzles Broad Tactical Training

Standard Puzzles give you access to over five million puzzles from the Lichess database.
You can filter by rating, themes, and openings to match your training goals, while tracking progress through an Elo based rating system and detailed session statistics. This mode is ideal for sharpening general tactical skills alongside your opening preparation.

Repertoire Puzzles Tactics From Your Own Openings

Repertoire Puzzles take tactical training one step further by generating puzzles directly from your own repertoires.
By selecting a repertoire, Chessboard Magic automatically finds tactical positions that arise from the openings you actually play. This tightly connects opening study with tactical execution, reinforcing patterns you are far more likely to encounter in real games.

Gap Analysis AI Driven Repertoire Improvements

The new Gap Analysis feature introduces a more modern, AI driven way of improving your opening repertoires.
Rather than relying purely on static theory or traditional engine tables, Gap Analysis allows you to analyse your repertoire using human centric and neural network engines such as Maia and Leela Chess Zero. These engines are especially good at identifying moves that are natural, practical, and commonly played, often reflecting how positions are handled in real games rather than purely theoretical best play.
By reviewing your existing lines through this lens, Gap Analysis can suggest possible improvements and additions. These are moves you may not currently include, but which AI models consider important for maintaining healthy and resilient positions. The goal is not theoretical perfection, but to help you build repertoires that feel more complete, more realistic, and better suited to practical play.
From the results, you can jump straight to the relevant position, review the suggestion in context, and decide whether or not to add it to your repertoire. It creates a natural bridge between modern AI evaluation and long term opening preparation.

Leela Chess Zero Analytics and Practice Integration

Leela Chess Zero has now been integrated into the platform in two complementary ways.
As an Analytics Card, Leela provides a neural network perspective that often highlights long term plans, positional ideas, and intuitive concepts that differ from traditional engine evaluations. This makes it especially useful for understanding why a position works, not just how to calculate it.
As a Practice Mode opponent, Leela brings a distinctive, pattern based playing style that feels very different from Stockfish. Training against Leela encourages deeper positional understanding and strategic awareness, and adds valuable variety to practice sessions.

MCP Enabled API Endpoints for AI Based Analysis

New API endpoints have been added to support Model Context Protocol (MCP) access.
These endpoints allow external AI assistants to securely read your repertoires and perform position analysis using accurate, real time data from your account. This opens the door to richer AI assisted study workflows, where an assistant can understand your opening choices, interpret positions in context, and support your training in a more meaningful way.
This infrastructure has been designed with future AI driven study tools in mind, while keeping access controlled and secure.

Five New Grandmaster Playbooks and Practice Bots

This release introduces five new Grandmaster Playbooks:

• Jan Timman
• Emanuel Lasker
• Adolf Anderssen
• Vassily Ivanchuk
• Wilhelm Steinitz

Each Playbook focuses on the player’s characteristic style, recurring themes, and instructive games. To deepen the experience, all five Grandmasters are also available as personality inspired practice bots, allowing you to train directly against move choices shaped by their strategic tendencies.
Studying a style and then immediately playing against it creates a strong feedback loop between understanding and application.

ChessDB Card Natural Language Commentary

The ChessDB Card has been significantly enhanced with natural language commentary.
Instead of relying only on engine evaluations, you can now read clear, human readable explanations that describe key ideas, threats, plans, and mistakes in the current position. A toggle lets you show or hide commentary whenever you want, and a subtle typewriter style animation makes longer explanations easier to read.
A new Add as Comment button allows you to attach the explanation directly to the current move in your repertoire, turning engine insight into permanent, personalised annotations.

A New Repertoire Builder Homepage

A new homepage has been added to the Repertoire Builder to better introduce the feature to new users.
It offers clearer explanations of what the Repertoire Builder does, how it fits into the broader Chessboard Magic workflow, and what to expect before getting started, making the onboarding experience smoother and more approachable.

Coming Soon

The next release will focus heavily on a major revamp of the Training Module.
Repertoire Training, Position Trainer, Weakness Trainer, Spaced Trainer, and Master Game Review will all receive visual improvements, additional functionality, and deeper integration with the rest of the platform. The goal is to make training feel more cohesive, more adaptive, and more closely connected to your repertoire work.
As the platform has grown to hundreds of features, I’m also planning a Simplified View that highlights the core tools. This will make onboarding easier and help new users get comfortable before diving into the more advanced functionality. Alongside this, I’ll be spending more time reducing the bug backlog and improving overall stability.
After eight months, the growth of Chessboard Magic has exceeded anything I imagined. We’ve reached 4,000 registered users, created just under 10,000 repertoires, and saved 1.4 million moves on the platform.
This has been quite a journey, and I’m incredibly grateful to everyone who has used the tool, shared feedback, and supported its development. I’m excited to continue building and to share what comes next with you all.

Kind regards,
Toan Hoang (@HollowLeaf)