Stockfish analysis. Local. Private.
Stockfish analysis. Local. Private.
Kiebitz is a local-first chess application available for Windows and Android. Its primary function is to import a user's complete game history from chess.com and Lichess, then analyze those games using a native Stockfish engine. The analysis runs entirely on the user's device, with no account creation, no cloud storage, and no telemetry. All data is stored locally in a SQLite file on the user's machine. The app provides several features for understanding playing patterns. The Insights section offers four evaluation pages that cover overall playing strength, openings separated by color, behavioral patterns such as time trouble and long sessions, and the phases where errors most frequently occur. The Repertoire feature builds a position tree from the user's own games and applies FSRS spaced repetition, similar to flashcard applications, to schedule reviews of specific moves. For training, Kiebitz includes an offline Lichess puzzle database, generates puzzles from positions where the user missed a winning move, and offers curated endgame drills against the engine, optionally using Syzygy tablebases. Additional capabilities include a drag-and-drop planner for weekly training, local reminders, and the ability to add moves from user games directly into the opening book. A typical workflow begins by entering a chess.com or Lichess username. The app fetches the entire game history, deduplicating when necessary, and can also import PGN files for over-the-board games. A native Stockfish engine then processes the queue in the background, providing per-move evaluations and classifying inaccuracies, mistakes, and blunders. The results are presented as aggregated statistics over months rather than individual game scores, helping users identify where their play deteriorates and under which conditions. Technically, Kiebitz is designed to operate without a server. Each device runs independently unless paired through QR code scanning, which is the only use of the camera. Users can train on a phone while on the move and analyze on a desktop in the evening, with no synchronization required. The app is currently in beta and available for free via GitHub Releases, with no account needed.