Waza 是一套面向工程实践的 Skill 集合，把工程师每天都会做的关键动作沉淀成可复用流程：先思考、再设计、做检查、系统排错、深度学习、可靠阅读、自然写作、配置体检。
/think：新功能或架构前，先做方案设计与风险验证。/design：做界面时给出明确审美方向，避免模板化 UI。/check：提交前做代码审查、风险分级和验证。/hunt：遇到 bug 时先找根因，再修复。/write：中英文写作润色，减少 AI 腔。/learn：从资料收集到成文发布的六阶段学习流程。/read：把网页或 PDF 拉取为干净的 Markdown。/health：审计 agent 配置健康度，定位失效层。
# Claude Code
npx skills add tw93/Waza -a claude-code -g -y
# 其他 agent（交互选择）
npx skills add tw93/Waza也可通过 .agents/skills/ 自动发现技能（取决于你的 agent 运行环境）。
/think、/hunt、/learn、/write、/design 可在多数 agent 直接使用（含 Codex）。平台相关能力会自动降级：例如缺少子代理接口时，/check 会回退到本地审查；缺少 AskUserQuestion 时，/design 与 /learn 会改为纯文本提问；/health 同时支持 .claude 与 .codex 路径，并识别 .agents/skills/。
Waza (技, わざ) is a Japanese martial arts term for technique: a move practiced until it becomes instinct.
A good engineer does not just write code. They think through requirements, review their own work, debug systematically, design interfaces that feel intentional, and read primary sources. They write clearly, and learn new domains by producing output, not consuming content.
AI makes you faster. It doesn't make you think more clearly, ship more carefully, or understand more deeply. Waza packages these habits into skills Claude can run.
Each engineering habit gets a Claude Code skill. Type the slash command, Claude follows the playbook.
| Skill | When | What it does |
|---|---|---|
/think |
Before building anything new | Challenges the problem, pressure-tests the design, validates architecture before any code is written. |
/design |
Building frontend interfaces | Produces distinctive UI with a committed aesthetic direction, not generic defaults. |
/check |
After a task, before merging | Reviews the diff, auto-fixes safe issues, blocks destructive commands, verifies with evidence. |
/hunt |
Any bug or unexpected behavior | Systematic debugging. Root cause confirmed before any fix is applied. |
/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 | Fetches content as clean Markdown via proxy cascade script. Dedicated handlers for WeChat and Feishu. |
/health |
Auditing Claude Code setup | Checks CLAUDE.md, rules, skills, hooks, MCP, and behavior. Flags issues by severity. |
Each skill is a folder, not just a markdown file. Skills include reference docs, helper scripts, and gotchas sections built from real project failures.
Statusline
A minimal Claude Code statusline that shows only what matters: context window usage, 5-hour quota, and 7-day quota, each with the time remaining until reset.
Color coding: green below 70%, yellow at 70-85%, red above 85% for context; blue, magenta, red for quota thresholds. No progress bars, no noise.
curl -sL https://raw.githubusercontent.com/tw93/Waza/main/scripts/setup-statusline.sh | bashEnglish Coaching
Working with AI in English just works better. The model thinks in English, most of the good resources are in English, and clear writing compounds over time.
Passive grammar correction on every reply. Claude flags mistakes with the pattern name so you learn why.
😇 it is not good to be read → it's hard to read (Unnatural phrasing)
curl -sL https://raw.githubusercontent.com/tw93/Waza/main/templates/coaching-en.md >> ~/.claude/CLAUDE.mdClaude Code:
npx skills add tw93/Waza -a claude-code -g -yOther agents: pick your agent interactively.
npx skills add tw93/WazaRequires Node 18+. Skills are also available at .agents/skills/ for agents that auto-discover from that path.
Install a single skill:
npx skills add tw93/Waza -a claude-code -s health -yReplace health with any skill name.
To protect against destructive git commands (git push -f, git checkout ., git clean -f), add them to the deny list in your agent settings file (~/.claude/settings.json or ~/.codex/settings.json).
Compatibility: Core instruction skills /think, /hunt, /learn, /write, /design run on any agent, including Codex. Platform-specific tool calls gracefully degrade: /check falls back to local specialist passes when sub-agent APIs are unavailable; /design and /learn use plain-text questions when AskUserQuestion is unavailable; /health supports both .claude and .codex paths and also discovers skills from .agents/skills/.
Tools like Superpowers and gstack are impressive, but they are heavy. Too many skills, too much configuration, too steep a learning curve for engineers who just want to get things done.
There's also a subtler problem. Every rule the author writes becomes a ceiling. The model can only do what the instructions say and can't go further. Waza goes the other direction. Each skill sets a clear goal and the constraints that matter, then steps back. As models improve, that restraint pays compound interest.
Eight skills for the habits that actually matter. Each does one thing, has a clear trigger, and stays out of the way. Not complete by design, just the right amount done well.
Built from patterns across real projects, refined through actual use. Every gotcha traces to a real failure: a wrong code path that took four rounds to find, a release posted before artifacts were uploaded, a server restarted eight times without reading the error. 30 days, 300+ sessions, 7 projects, 500 hours.
The /health skill is based on the six-layer framework described in this post.
- If Waza helped you, share it with friends or give it a star.
- Got ideas or bugs? Open an issue or PR, feel free to contribute your best AI model.
- I have two cats, TangYuan and Coke. If you think Waza delights your life, you can feed them canned food 🥩.
MIT License. Feel free to use Waza and contribute.