A personal chess coach, built from your games
Your next lesson is already on the board
Type your Chess.com or Lichess username. The coach reads your recent games, finds the mistakes that cost you points, and turns them into a short daily plan.
No account handy? Watch the coach read or , or browse the training rooms.
vs mira_endgames · blitz
move 4 · Ng5
The coach, mid-read
Example
Coach
Scanning
Reading the loss to mira_endgames at depth 22. Verdicts land as the scan reaches them.
Coach
Move 6 · brilliant
Book theory, the Fried Liver sacrifice: the knight goes so Qf3+ can pull the king forward. After Kxf7 Qf3+, Ke6 is the only way to save the d5 knight, so the king stays stuck in the center holding it.
Coach
Move 9 · mistake
Qe4 threatens nothing: the e5 pawn shields the king, so the queen just stands there while Black catches up. c6 props up the d5 knight and Black is already slightly better; Bb3, castling, or the d4 break were all better moves.
Coach
Move 17 · blunder
Nf4 hits the queen, and the retreat decides the game. Qf3 keeps a clear edge, covering the g4 and f3 squares the attack needs. Qd2 abandons the c4 bishop, whose only defender was the queen: Bxc4 wins it, and Nxg2 then Qg4+ rips the king open.This example runs on sample data; hover pauses it, the chart scrubs it. Your version is built from your own games and stays on your device.
Every game, reviewed free
Your games load in seconds. Open one and Stockfish walks through it right in your browser: a verdict on every move, a chart of where the game turned, and the coach explaining the turn in plain words. Chess.com sells this as a premium feature. Here it is the front door: no daily cap, no signup.
Your last games, graded
ExampleThe first game's report is on the right; every row opens one like it.
The first game’s report
move 17 · Qd2
One real brilliancy, then one blunder that undid it, in a whole-game read at depth 22. Click the chart; the real report drives a replay board.
This example runs on sample data. Your version is built from your own games and stays on your device.
The deep dive
Where your games actually go wrong
The deep dive is the slow pass. Stockfish replays your latest games move by move and re-checks every suspect moment at higher depth before it counts. The first run takes a while; that is the price of not guessing.
Out comes the shape of your game: how often you blunder, which phase leaks accuracy, what the clock does to your decisions, and which openings keep hurting you.
The shape of your game
ExampleBlunder control 34: about 1.6 confirmed blunders a game.
Openings 66: the first ten moves hold up.
Every axis is scored from verified moments, never from vibes. Unmeasured axes say so instead of guessing.
This example runs on sample data. Your version is built from your own games and stays on your device.
What the pass verified
Example12 games · scanned at depth 20 · every verdict re-checked at depth 22
Centipawns given away, per move
The blunders have a shape: 7 hung pieces, 9 winning positions let slip, 3 missed captures.
31% of confirmed mistakes came with under 20 seconds on the clock, against 14% of moves overall.
This example runs on sample data. Your version is built from your own games and stays on your device.
What tomorrow looks like
ExampleUp next · From your games
Repair the Qd2 blunder vs mira_endgames
The position comes back with the continuation hidden. Find the move that keeps the edge, and that blunder is fixed for good.
Tactics · 0/5
Hanging-piece puzzles
The blunder was a hanging piece; these five puzzles train exactly that.
Opening practice
Two Knights: the knight attack
You reach this line in most of your e4 games; the follow-up has to be automatic.
Sample plan. Yours is rebuilt each day from what the coach measured in your own games.
The review becomes tomorrow’s work
Most sites stop at the report. Here every confirmed mistake comes back as a position with the continuation hidden. Find the move you missed and it is fixed, retired from the queue for good. A backlog of seventy blunders is meant to shrink, not to become flashcards.
The daily plan pulls from that queue and from whatever else the evidence flags: the opening line you keep fumbling, the tactic you keep missing, the endgame that keeps slipping.
Behind it sits a growing library of annotated opening lines, positional exercises, adaptive puzzles, engine drills, and annotated classics.
