Extract root causes from multi-agent coding session logs.
npx skills add https://github.com/shoji9x9/skills --skill kaizen-shoji9x9
Please help me install this Agent Skill. Skill: kaizen Source: https://github.com/shoji9x9/skills/tree/main/.agents/skills/kaizen Command: npx skills add https://github.com/shoji9x9/skills --skill kaizen-shoji9x9
Coding agents often repeat the same mistakes across sessions because there is no systematic way to capture lessons learned from errors, analyze their root causes, and apply fixes to project rules, documentation, and hooks to prevent recurrence. This skill eliminates that gap by turning session failures into permanent, shared improvements for all team agents.
.agents/rules/, and ensures all future agents avoid the same mistake.Use the kaizen skill to extract lessons from your latest coding session and apply any pending improvements to your project's rules and documentation.
High-intent search queries and answers about installing and using this skill.
You can prevent repeat failures by scanning multi-agent session logs to extract errors and identify their root causes. The extracted lessons are then applied to project rules and hooks to permanently correct agent behavior.
Root cause analysis for coding agent session logs is the process of scanning multi-agent activity for failures and performing a minimum 3-layer "why" analysis. This identifies actionable root causes instead of just treating surface symptoms to prevent future errors.
Yes, the lesson extraction and application workflow is designed for software development using Claude Code, Codex, and GitHub Copilot. Extracted lessons are applied to shared project rules, ensuring all these agents benefit from past mistakes.
Yes, you can configure git hooks to block commits if unextracted session activity exists. The workflow uses sentinel files to mark unextracted sessions and injects pending lessons into agent context at session start to enforce the improvement cycle.
You apply extracted coding session lessons by integrating with a multiagent-setup skill that updates project rules, documentation, and custom skills. This ensures the root cause fixes are shared across all team agents and future sessions.
BumblebiberInstall and configure the hmem MCP server across system-wide or project-local scopes.
ExsyDevPersist project-specific AI rules to .ai-factory/RULES.md for loading before task execution.
AzureAutomate end-to-end testing of Azure Functions Skills across GitHub Copilot, Claude Code, and Codex.
glowingkittyTriage failing test runs into root-cause groups and apply targeted fixes.