dsh-go-sensei
Run the following command in DeepSeek Harness:
dsh plugin install Zhuang-A/dsh-go-sensei
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install Zhuang-A/dsh-go-sensei in the DSH terminal to install this plugin; the source repository is https://github.com/Zhuang-A/dsh-go-sensei .
About this plugin
After a game, what hurts most is not the result itself but not knowing exactly which move went wrong, why, and what concrete alternative would have been better. dsh-go-sensei hands your SGF game record to a Go-teacher persona inside DSH, walking through every problem move: its original intent, the flaw, and a specific better line with follow-up variations.
Follow-up questions come with annotated diagrams (moves numbered 1-9 / A-Z, key stones marked with triangles and other symbols); interactive win-rate and score-difference curves let you click any point to jump to that move; explanations can be written back into a -sensei.sgf copy alongside the original, or exported as a Markdown review report. On Windows the plugin ships with bundled KataGo v1.16.4 and 18b weights so missing analysis data is computed automatically with zero extra downloads; on other platforms or without an engine it degrades gracefully to pure-theory explanations.
Built for amateur players from 18k up to 1d. Explanation depth adapts to the weaker player's level listed in the game record, so beginners get everyday analogies while stronger players get direct strategic discussion. No Java, no separate Go software, no manual engine-path setup: drop an SGF file into a session and ask for a review to get started.
Screenshots
Use Cases
- Review a finished game move by move to spot big blunders and imprecisions
- Ask about an alternative for a specific move and see an annotated variation diagram
- Write explanations back into a -sensei SGF copy or export a Markdown review report
Best For
- Amateur players (18k-1d) seeking move-by-move coaching without a live teacher
- Players who want per-move win-rate and score trends visualized as interactive curves
- SGF-centric players who prefer explanations embedded directly in game-file comments
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.

