One command scores your project 0-100 for Claude Code optimization. Smart auto-fix generates framework-aware CLAUDE.md, hooks, and commands. Never overwrites your existing config.
Three steps to a fully optimized Claude Code project.
Scans your project against 63 best practices. Detects your framework, TypeScript, Docker, and more.
npx claudex-setup
Generates smart CLAUDE.md with Mermaid diagrams, hooks, commands, agents, and rules — all tailored to your stack.
npx claudex-setup setup
Monitors your project live. Re-audits on every change and shows score delta in real time.
npx claudex-setup watch
Every way you might want to optimize your Claude Code experience.
Score 0-100 against 63 checks. Detects framework, TypeScript, Docker. Shows critical, high, and quick wins.
npx claudex-setup
Generates CLAUDE.md that detects your scripts, creates Mermaid diagrams from your directory structure, adds framework-specific rules.
npx claudex-setup setup
Step-by-step guided tour through every improvement. Pick what to fix, skip what you don't need.
npx claudex-setup interactive
Live monitoring. Re-audits on every config change, shows score delta in real time. Keep it running while you work.
npx claudex-setup watch
Generate a shields.io badge for your README. Show the world your Claude Code readiness score.
npx claudex-setup badge
Claude reads your actual config and gives intelligent, specific feedback. Not pattern matching — real AI analysis. Requires API key.
npx claudex-setup deep-review
GitHub Action that audits every PR. Set a minimum score threshold. Fail the build if standards drop.
uses: DnaFin/claudex-setup@main
Not a generic template. Actually analyzes your project.
Reads package.json and includes your actual test, build, lint, and dev commands.
Next.js Server Components, Django models, FastAPI Pydantic, React hooks — each gets specific guidelines.
Detects strict mode, adds TS-specific rules. No @ts-ignore without tracking.
Scans your directories and generates an architecture diagram. More token-efficient than prose (per Anthropic docs).
Adds <constraints> and <verification> blocks with context-aware rules. 30% quality improvement.
Auto-generates a checklist based on your actual test, lint, and build commands.
Every check is backed by tested research from the CLAUDEX catalog.
Project instructions Claude reads every session
Test/lint commands so Claude checks its own work
Block .env reads, deny rules for sensitive files
PreToolUse + PostToolUse automation
Architecture visualization. More token-efficient than prose
Structured prompts. 30% quality boost
Custom /test, /deploy, /review, /fix
Built-in OWASP Top 10 scanning
Security reviewer, test writer subagents
Path-specific conventions per file type
On-demand domain knowledge
Context7, database, and API integrations
Audit is read-only — it never touches your files. Setup only creates new files in .claude/ and a CLAUDE.md. It never modifies existing source code. Everything runs locally on your machine — no data is sent anywhere.
/init?Claude Code's /init generates a basic CLAUDE.md. We go much further: 63 audit checks, smart CLAUDE.md with Mermaid diagrams, PreToolUse/PostToolUse hooks, security agents, framework-specific rules, XML constraint blocks, and verification criteria — all backed by research from 1,107 cataloged Claude Code entries.
Based on our research: Mermaid diagrams are more token-efficient than prose descriptions (per Anthropic docs). Hooks enforce rules at 100% (vs ~80% for CLAUDE.md instructions alone). Verification loops are Anthropic's #1 recommended best practice. The impact depends on your project and workflow.
The audit is Claude Code-specific. However, CLAUDE.md files are recognized by many AI coding tools. The skills format was adopted as a cross-platform standard by OpenAI Codex CLI in December 2025. Hooks and agents are Claude Code-only features.
From a systematic research project covering all 73 official Claude Code documentation pages, 100+ community tools and MCP servers verified via GitHub API, Anthropic blog posts, benchmark papers, and hands-on experiments. The count includes features, techniques, patterns, tools, stats, and limitations — not all are actionable. 954 were verified with real evidence. The research is ongoing.
Audit will still show you what else you might be missing (hooks, agents, rules, MCP config). Setup will never overwrite your existing files — it only creates what's missing. If your project is already well-configured, the tool will tell you. A hand-crafted CLAUDE.md that reflects your real conventions will always be better than a generated one.
63 checks. Read-only audit. Never overwrites your existing setup. If you already have a good config, it'll tell you.