Waza README

The project's README from github.com/tw93/Waza. As one markdown file: README.md.

Waza

Engineering habits you already know, turned into skills AI agents can run.

LLM guide

Tests Stars Version License Twitter

Skills

Each engineering habit gets an installed skill. In Claude Code, type the slash command. In Codex, invoke the installed skill by name and follow the same playbook.

Skill When What it does
/think Before building anything new Challenges the problem, pressure-tests the design, and produces a decision-complete plan another agent can implement.
/ui Building frontend interfaces Produces distinctive UI, including screenshot-driven aesthetic iteration, with a committed direction rather than generic defaults.
/check After a task, before merging or release Reviews the diff against project constraints, verifies the result, and handles approved release and maintainer actions.
/hunt Any bug, regression, or unexpected behavior Systematic debugging. Root cause confirmed before any fix is applied, especially when something used to work.
/write Writing or editing prose Rewrites prose to sound natural in Chinese and English. Cuts stiff, formulaic phrasing.
/learn Diving into an unfamiliar domain Six-phase research workflow: collect, digest, outline, fill in, refine, then self-review and publish.
/read Any URL or PDF Returns a concise summary, or clean Markdown for conversion, quotes, citations, saving, or further research.
/health Auditing Agent Health Checks Codex, Claude Code, project instructions, verifier output, and AI maintainability with a budget-aware summary pass before deep inspection.

Each skill is a folder with reference docs, helper scripts, and gotchas from real failures.

Install

Claude Code, Codex, Cursor, and other agents

npx skills add tw93/Waza -a claude-code codex cursor -g -y

One copy lands in ~/.agents/skills, the shared skills directory. Claude Code is symlinked in; Codex, Cursor, Gemini CLI, Copilot, Amp, Kimi Code CLI, and every other agent that reads that directory picks the eight skills up as /check, /think, and so on. Agents with a private skills directory take their id after -a (for example antigravity-cli or qwen-code). Update with npx skills update -g -y.

Host plugin, if you prefer the host's own update command (skills are namespaced, /waza:check)

# Claude Code (update: claude plugin update waza)
/plugin marketplace add tw93/Waza
/plugin install waza@waza

# Codex (update: codex plugin marketplace upgrade waza, then codex plugin add waza@waza)
codex plugin marketplace add tw93/Waza
codex plugin add waza@waza

Claude Desktop: download waza.zip, open Customize > Skills > "+" > Create skill, and upload the ZIP. To update, click "..." on the skill card, choose Replace, and upload the latest ZIP.

Pi: pi install npm:@tw93/waza, update with pi update npm:@tw93/waza.

Chaining Skills

You decide how skills chain together. Each skill stops at the requested outcome; an explicitly authorized workflow continues without asking you to approve every transition.

Common workflows:

Project Context

Waza ships only generic engineering habits. /check becomes project-aware at runtime by reading the target repository's public context (READMEs, package manifests, Makefiles, CI workflows) and your task constraints, never private paths, credentials, or tokens. See skills/check/references/project-context.md for the review context template.

Extras

Statusline

A minimal statusline for Claude Code: context window, 5-hour quota, and 7-day quota. Color-coded by usage, no progress bars, no noise.

(
  set -e
  WAZA_STATUSLINE_SCRIPT="$(mktemp -t waza-statusline.XXXXXX)"
  trap 'rm -f "$WAZA_STATUSLINE_SCRIPT"' EXIT
  curl -fL https://github.com/tw93/Waza/releases/latest/download/setup-statusline.sh -o "$WAZA_STATUSLINE_SCRIPT"
  # review it first: less "$WAZA_STATUSLINE_SCRIPT"
  bash "$WAZA_STATUSLINE_SCRIPT"
)

Codex has native statusline items. Add to ~/.codex/config.toml:

[tui]
status_line = ["model-with-reasoning", "current-dir", "context-used", "five-hour-limit", "weekly-limit"]
status_line_use_colors = true

Codex shows remaining quota; the Claude Code statusline above shows used percentage (upstream does not yet expose five-hour-used / weekly-used).

Optional Rules

Three independent toggles. Copy the ones you want (swap claude-code for codex or antigravity-cli on those agents):

(
  set -e
  WAZA_RULE_SCRIPT="$(mktemp -t waza-rule.XXXXXX)"
  trap 'rm -f "$WAZA_RULE_SCRIPT"' EXIT
  curl -fL https://github.com/tw93/Waza/releases/latest/download/setup-rule.sh -o "$WAZA_RULE_SCRIPT"
  # review it first: less "$WAZA_RULE_SCRIPT"

  # English coaching: appends a short 😇 correction when your prompt has an English mistake
  bash "$WAZA_RULE_SCRIPT" english claude-code

  # Anti-patterns: always-on cross-skill guardrails (read before acting, no scope creep, no unsolicited summaries)
  bash "$WAZA_RULE_SCRIPT" anti-patterns claude-code

  # Routing hint: tells non-Claude hosts to prefer Waza skills when a request matches their triggers
  bash "$WAZA_RULE_SCRIPT" waza-routing claude-code
)

Curl URLs use the latest GitHub release asset. Set WAZA_REF=main before the command if you want bleeding-edge scripts.

Why

Waza (技, わざ) is a Japanese martial arts term for technique: a move practiced until it becomes instinct.

A good engineer does more than write code. They pressure-test requirements, debug to root cause, review their own diffs, and read primary sources. AI has the raw output for all of it, but without structure that output drifts into generic, imprecise work. Each Waza skill states the outcome, the red lines, and how the result gets verified, then steps back and lets the model choose the path. As models improve, that restraint pays compound interest.

Tools like Superpowers and gstack are powerful but heavy: too many skills, too much configuration. Waza stays small, eight skills for the habits that actually matter, each with one job and a clear trigger. Built from real projects and refined through 300+ sessions across 7 projects, every gotcha traces to a real failure. The /health skill grew from the six-layer Claude Code framework in this post.

Part of a trilogy: Kaku (書く) writes code, Waza (技) drills habits, Kami (紙) ships documents. Think of them as a family: Kaku is the dad, Waza the big sister, Kami the little sister.

Uninstall

npx skills remove tw93/Waza -g
rm -f ~/.claude/statusline.sh
rm -f ~/.claude/rules/english.md
rm -f ~/.claude/rules/anti-patterns.md
rm -f ~/.claude/rules/waza-routing.md

For Claude Desktop, delete Waza from Customize > Skills. For Codex rule installs, remove the marked Waza blocks from ~/.codex/AGENTS.md.

Support

These lovely people already did 🐱

License

MIT License. Feel free to use Waza and contribute.

Waza (tw93/Waza)

claude-skill
Searching Database…

7.3K ·GitHub
Waza
What is this?Waza is a pack of eight 'skills' that turn proven software engineering habits (planning, UI design, code review, debugging, writing, research, reading, agent health checks) into slash commands that coding agents like Claude Code can run.
Is this a library, or something else?It is not code you import — it is a shared skills directory. Each of the eight skills is a folder with reference docs, helper scripts and known gotchas that your AI agent reads and executes as a slash command.
How do I install the skills?Run `npx skills update -g -y` to drop a shared copy into `~/.agents/skills`; for Claude Desktop, download waza.zip and add it via Customize > Skills > Create skill.
Which coding agents can actually use these skills?Claude Code is symlinked in automatically; Codex, Cursor, Gemini CLI, Copilot, Amp, Kimi Code CLI, Pi, and any other agent that reads the shared skills directory pick up the same eight skills.
Does this know anything about my specific project?No — Waza only ships generic engineering habits. The /check skill becomes project-aware at runtime by reading a repo's public context (READMEs, manifests, CI files), but it never touches private paths, credentials, or tokens.
What are the eight skills included?/think (pressure-test a plan before building), /ui (distinctive frontend design), /check (diff review before merge/release), /hunt (root-cause debugging), /write (natural prose editing), /learn (six-phase research workflow), /read (summarize URLs/PDFs), and /health (audit agent setup health).
Can I chain skills into a workflow?Yes. Each skill stops at its outcome by default, but you can authorize a chain like /think → implement → /check → merge, or /hunt → fix → /check, and the agent continues without asking for approval at every step.
Why build Waza instead of using something like Superpowers?The author found tools like Superpowers and gstack powerful but too heavy, with too many skills and too much configuration. Waza deliberately stays small: eight skills, each with one job and a clear trigger, refined through 300+ real sessions.
What is the statusline extra for?It's a minimal Claude Code statusline showing context window usage and 5-hour/7-day quota, color-coded with no progress bars; Codex instead gets native statusline config via ~/.codex/config.toml.
What are similar projects?superpowers-zh, open-steps, Coding-Agent-Skills, pstack, pstack-claude, superpowers, evondevKit, brag
Which topics is it in?npx, claude-skill, copilot, gemini-cli, claude-code-skill, code-review, debugging, workflow, design, cursor, skills, agents
Creator
tw93
Size
6.1 MB
Stars
7.3K
Updated
CommitOct 18Releasev3.40.0Oct 15
Age
7mo
Commits · 90 Days