claudeignore

Generate optimized .claudeignore files to reduce token usage and costs in AI coding assistants.

What it does

Creates a .claudeignore file in your project root to prevent Claude Code from loading unnecessary files (node_modules, build artifacts, etc.). This can reduce token consumption by 90%+ in large projects.

Usage

/claudeignore [template]

Examples:

• /claudeignore - Auto-detect project type and generate
• /claudeignore nextjs - Generate for Next.js project
• /claudeignore python - Generate for Python project
• /claudeignore go - Generate for Go project

Available Templates

• nextjs - Next.js/React projects
• react - React (CRA, Vite) projects
• vue - Vue.js projects
• python - Python projects with pip/poetry
• go - Go projects
• rust - Rust projects
• node - Generic Node.js projects
• full - Comprehensive template (all common patterns)

Step 1: Detect Project Type

First, analyze the project structure to determine the framework/language:

Look for key indicators:

• package.json → Check for next/react/vue in dependencies
• requirements.txt or pyproject.toml → Python
• go.mod → Go
• Cargo.toml → Rust

Step 2: Choose Base Template

Based on detection, select the appropriate template. If user specified a template, use that instead.

Step 3: Generate .claudeignore

Create the .claudeignore file with the selected template. Use the templates below:

Templates

Next.js / React

# Dependencies
node_modules/
.pnp/
.pnp.js

# Build outputs
.next/
out/
build/
dist/

# Cache
.cache/
.turbo/
.swc/

# Testing
coverage/

# Environment
.env*.local

# Logs
*.log
logs/

# OS
.DS_Store
Thumbs.db

# IDE
.vscode/
.idea/
*.swp
*.swo

# Package manager
.yarn/
.pnpm-store/
package-lock.json
yarn.lock
pnpm-lock.yaml

# Media & Static Assets
*.png
*.jpg
*.jpeg
*.gif
*.svg
*.ico
*.webp
*.woff
*.woff2
*.ttf
*.eot
*.mp4
*.mp3
*.webm
public/images/
public/videos/
assets/images/

Python

# Virtual environments
venv/
env/
ENV/
.venv/

# Python cache
__pycache__/
*.py[cod]
*$py.class
.pytest_cache/

# Distribution
dist/
build/
*.egg-info/
.eggs/

# Testing
.coverage
htmlcov/
.tox/

# Jupyter
.ipynb_checkpoints/
*.ipynb

# Environment
.env
.env.local

# Logs
*.log

# IDE
.vscode/
.idea/
*.swp

# OS
.DS_Store

# Media & Static Assets
*.png
*.jpg
*.jpeg
*.gif
*.svg
*.ico
*.webp
static/images/
media/

Go

# Binaries
*.exe
*.exe~
*.dll
*.so
*.dylib
bin/

# Test coverage
*.out
coverage.txt

# Go workspace
go.work
go.work.sum

# Vendor
vendor/

# IDE
.vscode/
.idea/
*.swp

# OS
.DS_Store

# Logs
*.log

# Media & Static Assets
*.png
*.jpg
*.jpeg
*.gif
*.svg
*.ico
*.webp
static/
public/

Rust

# Build outputs
target/
Cargo.lock

# Backup files
**/*.rs.bk

# IDE
.vscode/
.idea/
*.swp

# OS
.DS_Store

# Logs
*.log

# Media & Static Assets
*.png
*.jpg
*.jpeg
*.gif
*.svg
*.ico
*.webp
static/
assets/

Generic Node.js

# Dependencies
node_modules/

# Build
dist/
build/

# Logs
*.log
logs/

# Environment
.env
.env*.local

# OS
.DS_Store
Thumbs.db

# IDE
.vscode/
.idea/
*.swp

# Media & Static Assets
*.png
*.jpg
*.jpeg
*.gif
*.svg
*.ico
*.webp
*.woff
*.woff2
*.ttf
public/
static/
assets/images/

Full (Comprehensive)

# Dependencies
node_modules/
vendor/
.pnp/

# Build outputs
dist/
build/
out/
.next/
target/
*.egg-info/

# Cache
.cache/
.turbo/
.swc/
__pycache__/

# Virtual environments
venv/
.venv/
env/

# Package manager locks
package-lock.json
yarn.lock
pnpm-lock.yaml
Cargo.lock

# Testing
coverage/
.coverage
.pytest_cache/
*.test
*.out

# Logs
*.log
logs/

# Environment
.env
.env*.local

# IDE
.vscode/
.idea/
*.swp
*.swo

# OS
.DS_Store
Thumbs.db

# Git
.git/

# Media & Static Assets
*.png
*.jpg
*.jpeg
*.gif
*.svg
*.ico
*.webp
*.woff
*.woff2
*.ttf
*.eot
*.mp4
*.mp3
*.webm
*.pdf
public/
static/
assets/images/
media/
uploads/

Step 4: Calculate Savings

After creating the file, estimate token savings by checking sizes of major ignored directories:

Calculate percentage based on directory sizes. Present results:

✅ Created .claudeignore with [template] template

📊 Estimated Impact:
- Ignored directories: ~XXX MB (node_modules, build outputs, etc.)
- Total project size: ~XXX MB
- Estimated exclusion: ~XX% of project
- Token savings: ~90-95% per conversation
- Cost savings: ~$15-20 per 1M tokens

💡 Tip: Run 'claude --refresh' to apply changes to current session.

Important: Never use find . -type f as it can timeout on large projects with millions of files. Always use du -sh on specific directories instead.

Step 5: Offer Customization

Ask if user wants to:

• Add custom patterns
• Exclude specific directories
• See what files will still be included

Example Output

✅ Created .claudeignore for Next.js project

📊 Impact:
- 15,847 files excluded (node_modules, .next, etc.)
- ~450 MB removed from context
- Estimated 92% token reduction
- Save ~$18 per 1M tokens

💡 Your conversations will now be faster and cheaper!

Would you like to:
1. Add custom ignore patterns
2. Review what's still included
3. Done

Notes

• Always create in project root (where package.json/go.mod/etc. exists)
• If .claudeignore exists, ask before overwriting
• Suggest reviewing with cat .claudeignore after creation
• For mono-repos, consider multiple .claudeignore files in subdirectories

© License & Terms of Use Created by [@ziChen637523]. Licensed under CC BY-NC 4.0 (Creative Commons Attribution-NonCommercial).

✅ Free to use & share for personal and open-source projects. ❌ Strictly prohibited from being sold, packaged into paid courses/tools, or used for any commercial monetization without explicit written consent.