yo-measure

Insert symmetrical debug logs into rshogi and YaneuraOu to analyze search node count discrepancies.

15|2|Updated May 10, 2025
One-click install
npx skills add https://github.com/SH11235/rshogi --skill yo-measure
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: yo-measure
Source: https://github.com/SH11235/rshogi/tree/main/.claude/skills/yo-measure
Command: npx skills add https://github.com/SH11235/rshogi --skill yo-measure

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill pinpoints the exact cause of discrepancies in search node counts between the rshogi and YaneuraOu Shogi engines, enabling precise debugging and performance optimization.

Core Features & Use Cases

  • Deep Debugging: Inserts detailed logging into both engines to trace execution flow, variable states, and decision-making at a granular level.
  • Comparative Analysis: Facilitates side-by-side comparison of engine behavior, identifying subtle differences in search algorithms, pruning techniques, and evaluation functions.
  • Use Case: When rshogi and YaneuraOu produce different search node counts for the same position, this Skill helps you identify whether the difference stems from move ordering, pruning, transposition table usage, or evaluation function nuances.

Quick Start

Use the yo-measure skill to compare node counts for rshogi and YaneuraOu starting from the 'startpos' with a depth of 14, focusing on root move breakdowns.

Frequently Asked Questions about yo-measure

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I debug Shogi engine search node count discrepancies between rshogi and YaneuraOu?

Debug Shogi engine search node count discrepancies by inserting detailed, symmetrical debug logs into both rshogi and YaneuraOu codebases to trace execution flow and compare variable states. You achieve precise localization of root causes through step-by-step drill-downs and comparative analysis of execution traces.

What causes different search node counts in Shogi engines for the same position?

Different search node counts in Shogi engines stem from subtle differences in move ordering, pruning techniques, transposition table usage, or evaluation function nuances. Comparative analysis of execution traces identifies whether these algorithmic variations cause the performance discrepancies.

How to compare rshogi and YaneuraOu node counts starting from startpos?

Compare rshogi and YaneuraOu node counts by running the yo-measure skill from the 'startpos' with a depth of 14, focusing on root move breakdowns. This side-by-side comparison identifies exact algorithmic divergences in search behavior.

Can I toggle specific Shogi engine features to isolate search algorithm differences?

Yes, you can toggle specific Shogi engine features to isolate search algorithm differences. Feature toggling allows you to disable specific pruning techniques or evaluation functions, enabling precise step-by-step drill-downs to achieve perfect alignment in engine behavior.

What is the best way to achieve perfect alignment in Shogi engine behavior?

Achieve perfect alignment in Shogi engine behavior by inserting detailed logging into both codebases and performing comparative analysis of execution traces. Feature toggling and step-by-step drill-downs ultimately aim to match rshogi and YaneuraOu search node counts exactly.