Week 1 with Claude Code, I typed: "improve this project."
Claude read 4 files, edited 2, broke 1. 34,000 tokens. Nothing was fixed. I typed
/cost. $0.09 for zero work.
After a month of figuring out what actually works, the same changes now ship with 60–70% fewer tokens. This repo is the condensed version.
- "improve this project" → 34,000 tokens
+ "add RFC 5322 validation to validateEmail()
+ at src/auth/login.ts:87" → 4,000 tokensSame outcome. One-eighth the cost.
# Trims almost nothing
/compact
# Cuts 70% of context, quality unchanged
/compact Keep only changed functions and relevant test outputIn one long session this took me from 82,000 → 11,000 tokens.
node_modules/
dist/
*.log
coverage/
*.lockSix lines. File reads drop 3× to 10× depending on the project.
|
82 ready-made slash commands.
Copy to |
Copy-paste prompt shapes for bug fixes, refactors, reviews, migrations. |
Framework-ready memory files for Next 16, React 19, Django 5.2, Go 1.26, Rust. |
|
Drop-in |
Shortcuts, CLAUDE.md, permission modes, MCP, subagents — all in one file. |
Bookmark-worthy one-pager. Everything you'll reach for twice a day. |
82 skills across categories — click to fold
| Category | Count | Examples |
|---|---|---|
| Git & GitHub | 13 | /fix-issue · /pr-review · /commit · /release-notes |
| Testing | 7 | /test-generate · /test-coverage · /test-e2e |
| Frontend | 9 | /component-gen · /seo-check · /state-audit |
| Backend | 10 | /api-endpoint · /auth-middleware · /rate-limiter |
| Mobile | 4 | /react-native-component · /mobile-perf · /expo-setup |
| Security | 3 | /security-audit · /secret-scan · /dependency-audit |
| Database | 6 | /migration-gen · /query-optimize · /schema-review |
| + more | ... | See all → |
flowchart LR
A["Vague prompt"] -.->|Wastes tokens| X(["Fail"])
B["CLAUDE.md<br/>+ .claudeignore"] --> C["Specific prompt"]
C --> D["Run a skill"]
D -->|Long session| E["compact with hint"]
E --> F(["Ship"])
D --> F
style X fill:#ff6b6b,stroke:#000,color:#fff
style F fill:#51cf66,stroke:#000,color:#000
style B fill:#7B4FFF,stroke:#000,color:#fff
One command (installs core skills, .claudeignore, and CLAUDE.md scaffold):
curl -fsSL https://raw.githubusercontent.com/berkcangumusisik/claude-code-practices/main/install.sh | bashFlags: --all for every skill, --skills a,b,c to pick.
Or manually:
git clone https://github.com/berkcangumusisik/claude-code-practices.git
cp -r claude-code-practices/skills ./.claude/skills
claude
> /cost-audit| If you're... | Start here |
|---|---|
| New to Claude Code | Cheatsheet |
| Paying too much in tokens | Run /cost-audit on your session |
| Rewriting the same prompts | Prompt templates · Skill Library |
| Starting a new project | CLAUDE.md gallery |
| Tired of running formatters by hand | Hook recipes |
| Looking for more resources | Awesome Claude Code |
Does this work with Claude 4.6 / 4.7?
Yes. Every skill and pattern here is model-agnostic — they're about how you prompt Claude Code, not which model it's pointed at.
Do I need to copy all 82 skills?
No. Start with 3–5 you'd actually use this week. Add more as the pain points show up.
Will .claudeignore break anything?
No. It only controls what Claude reads — your build and editor are unaffected.
How do I measure the savings?
Run /cost before and after a session. The numbers in this README are from my own /cost outputs, not estimates.
Can I use this for a team?
Yes. Commit .claude/skills/ and a CLAUDE.md to your repo and the whole team picks them up automatically.
If this repo saved you tokens or hours, there are three ways to give back:
- Star the repo so the next person finds it
- Sponsor on GitHub to keep new skills and templates coming
- Open a PR with your own patterns (see below)
Found a better pattern, wrote a skill, spotted a bug? PRs welcome. See CONTRIBUTING.md.
If this saved you tokens, drop a star. That's how the next person finds it.
Built by @berkcangumusisik · MIT License