KROU4/claude-kaizen
Continuous improvement for Claude Code: measures how you really work, lets you pick fixes in an interactive console, then configures hooks, skills, agents and rules for you. /insights that acts.
ARCHETYPE
Pragmatist
A balanced, no-drama setup: some rules, some tools, nothing extreme.
Copy this rig
# review before running: this installs third-party code $ npx degit KROU4/claude-kaizen/agents ./rig-claude-kaizen/agents $ npx degit KROU4/claude-kaizen/skills ./rig-claude-kaizen/skills $ npx degit KROU4/claude-kaizen/hooks ./rig-claude-kaizen/hooks
MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit KROU4/claude-kaizen/skills/apply .claude/skills/apply $ npx degit KROU4/claude-kaizen/skills/measure .claude/skills/measure $ npx degit KROU4/claude-kaizen/skills/recipes .claude/skills/recipes $ npx degit KROU4/claude-kaizen/skills/start .claude/skills/start $ npx degit KROU4/claude-kaizen/skills/status .claude/skills/status $ npx degit KROU4/claude-kaizen/skills/undo .claude/skills/undo
$ curl -fsSL --create-dirs -o .claude/agents/kaizen-analyst.md https://raw.githubusercontent.com/KROU4/claude-kaizen/main/agents/kaizen-analyst.md $ curl -fsSL --create-dirs -o .claude/agents/kaizen-smith.md https://raw.githubusercontent.com/KROU4/claude-kaizen/main/agents/kaizen-smith.md $ curl -fsSL --create-dirs -o .claude/agents/kaizen-verifier.md https://raw.githubusercontent.com/KROU4/claude-kaizen/main/agents/kaizen-verifier.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} Skills (6)
Subagents (3)
| kaizen-analyst model: inherit | Reads the Kaizen measurement digest and drafts the qualitative part of the recommendations — a short workflow profile, concrete CLAUDE.md rules distilled from the user's repeated corrections, and skil |
| kaizen-smith model: inherit | Writes the content of a custom Kaizen recipe when a user picked one that needs authored text — a skill from a repeated prompt, CLAUDE.md rules from corrections, or a subagent definition. Produces file |
| kaizen-verifier model: haiku | Verifies what Kaizen just applied — every installed hook passes its self-test, every touched settings.json parses and contains the expected entries (and no longer contains the hooks a recipe removed), |
Similar rigs
archi-ai-labs/trim-kit
A small toolkit for setting up and maintaining Claude Code configuration per project. No hooks, no agents, no MCP - plain skills only.
Pragmatist 307 tok · · Q0
SAMARTHJAISWAL/aavedan
Finds jobs in India, remote and worldwide, ranks them against your resume, fills each application in Chrome, checks every answer, and sends only what you approve. Claude API + Playwright.
Pragmatist 976 tok · · Q0
NovaSeth/Landfall
Claude Code plugin that sizes the process to each coding task and checks the outcome in code: a task is done only when its gates pass again.
Pragmatist 787 tok · · Q0
sweetcornna/bunsen
开源 AI 科研工作台:让 Claude Code、Codex 等 Agent 在你自己的 GPU 服务器上做研究。运行记录、结论账本、产物画廊、显卡监控,断线不停。Self-hosted AI research workbench.
Pragmatist 353 tok · · Q0