awesome-claude-code-toolkit
Comprehensive toolkit for Claude Code with 135+ agents, 176+ plugins, 42 commands, 20 hooks, 15 rules, and 400K+ skills via SkillKit
How do I install this agent skill?
npx skills add https://github.com/reason-machines/claude-code-skills --skill awesome-claude-code-toolkitIs this agent skill safe to install?
- Gen Agent Trust Hubfail
The skill promotes a high-risk installation method where a shell script is downloaded from a third-party GitHub repository and executed directly via a pipe to bash. This allows for arbitrary code execution from unverified sources. The skill also facilitates numerous other downloads of scripts and plugins from multiple unverified third-party repositories.
- Socketwarn
1 alert: gptAnomaly
- Snykfail
Risk: CRITICAL · 4 issues
What does this agent skill do?
Awesome Claude Code Toolkit
Skill by ara.so — Claude Code Skills collection.
The most comprehensive toolkit for Claude Code, providing 135 agents, 35 curated skills (+400,000 via SkillKit), 42 commands, 176+ plugins, 20 hooks, 15 rules, 7 templates, 14 MCP configs, 26 companion apps, and 52+ ecosystem entries.
Installation
Via Plugin Marketplace (Recommended)
/plugin marketplace add rohitg00/awesome-claude-code-toolkit
Manual Clone
git clone https://github.com/rohitg00/awesome-claude-code-toolkit.git ~/.claude/plugins/claude-code-toolkit
One-Liner Install
curl -fsSL https://raw.githubusercontent.com/rohitg00/awesome-claude-code-toolkit/main/setup/install.sh | bash
Core Components
Plugins (176+)
The toolkit includes production-ready plugins across multiple domains:
Featured Plugins
pro-workflow - Battle-tested workflows with self-correcting memory:
/plugin marketplace add rohitg00/pro-workflow
gstack - Garry Tan's opinionated Claude Code setup:
/plugin marketplace add garrytan/gstack
agento-patronum - Security layer for sensitive files:
git clone https://github.com/emaarco/agento-patronum.git ~/.claude/plugins/agento-patronum
Domain-Specific Plugins
AWS Cost Optimization:
/plugin marketplace add prajapatimehul/aws-cost-saver
Code Quality & Reviews:
# Brooks-lint: Classic engineering book principles
/plugin marketplace add hyhmrright/brooks-lint
# Bouncer: Independent quality gate with Gemini
/plugin marketplace add buildingopen/bouncer
Project Management:
# CCPM: GitHub Issues + Git worktrees
/plugin marketplace add automazeio/ccpm
# Fractal: Recursive project management
/plugin marketplace add rmolines/fractal
Cost Control:
# Cost optimizer skill
npx skills add Sagargupta16/claude-cost-optimizer
# Usage analyzer
npm install -g ccusage
ccusage analyze
Agents (135)
The toolkit provides specialized agents for different roles:
Software Development Lifecycle
great_cto - Full SDLC pipeline with 7 agents:
// Agents included:
// - tech-lead: Architecture decisions
// - senior-dev: Code implementation
// - qa-engineer: Testing strategy
// - security-officer: Security audits
// - devops: CI/CD and infrastructure
// - l3-support: Production issues
// - project-auditor: Compliance checks
// Usage via plugin
/plugin marketplace add avelikiy/great_cto
Custom Agent Configuration
# agents/backend-architect.yml
name: backend-architect
role: Backend Service Architect
capabilities:
- API design with OpenAPI specs
- Database schema design
- Microservice decomposition
- Performance optimization
context:
- Always consider scalability
- Follow REST/GraphQL best practices
- Include authentication patterns
Skills (35 Curated + 400K via SkillKit)
Installing Skills
# Add individual skill
npx skills add <username>/<skill-name>
# Cost optimizer with templates
npx skills add Sagargupta16/claude-cost-optimizer
# Search SkillKit marketplace
npx skills search "database migration"
Creating Custom Skills
---
name: my-custom-skill
description: Custom skill for specific workflow
triggers:
- perform custom workflow
- run custom task
---
# My Custom Skill
## Usage
When user requests this skill:
1. Analyze requirements
2. Execute steps
3. Validate results
Commands (42)
Built-in Toolkit Commands
# Plugin management
/plugin marketplace search <query>
/plugin marketplace add <owner>/<repo>
/plugin list
/plugin remove <name>
# Skill management
/skill add <name>
/skill list
/skill remove <name>
# Agent management
/agent activate <name>
/agent list
/agent deactivate <name>
# Project scaffolding
npx claude-scaffold init
npx claude-scaffold update --all
# Cost analysis
ccusage analyze
ccusage report --monthly
ccusage diagnose --top 10
Skills Janitor Commands
# Audit skills
/skills-audit
# Deduplicate
/skills-dedupe
# Check validity
/skills-check
# Fix common issues
/skills-fix
# Usage tracking
/skills-usage
Hooks (20)
Hooks execute scripts at key lifecycle points:
Hook Types
// pre-approve.js - Validation before approval
module.exports = async (context) => {
const { files, command } = context;
// Block sensitive file access
const sensitivePatterns = ['.env', 'id_rsa', 'credentials'];
const blocked = files.filter(f =>
sensitivePatterns.some(p => f.includes(p))
);
if (blocked.length > 0) {
throw new Error(`Blocked access to: ${blocked.join(', ')}`);
}
};
// post-task.js - Archive after completion
module.exports = async (context) => {
const { task, result, metadata } = context;
// Log to memory system
await context.memory.store({
task: task.description,
result: result.summary,
timestamp: Date.now(),
cost: metadata.tokens
});
};
Agento Patronum Security Hooks
# Install security hooks
git clone https://github.com/emaarco/agento-patronum.git ~/.claude/hooks/security
# Configure protected patterns
cat > ~/.claude/hooks/security/config.json << 'EOF'
{
"patterns": {
"credentials": [".env", ".env.*", "*.pem", "id_rsa*"],
"cloud": ["~/.aws/*", "~/.kube/config"],
"secrets": ["secret*", "password*", "token*"]
}
}
EOF
Rules (15)
Rules guide Claude Code behavior:
# rules/cost-optimization.md
## Cost Optimization Rules
1. **Minimize Context**: Only include relevant files
2. **Cache Prompts**: Use consistent system prompts
3. **Batch Operations**: Group similar tasks
4. **Incremental Changes**: Avoid full rewrites
5. **Smart Diffs**: Use targeted edits
## Implementation
- Activate cost-mode skill for 30-60% savings
- Use Haiku for routine tasks
- Reserve Opus for complex architecture
# rules/security-first.md
## Security Rules
1. Never commit secrets to version control
2. Use environment variables: `process.env.API_KEY`
3. Validate all user input
4. Apply principle of least privilege
5. Enable audit logging for sensitive operations
Templates (7)
Project Templates
# Initialize with template
/template apply node-api
# Available templates:
# - node-api: Express REST API
# - react-app: React + TypeScript
# - python-cli: Python CLI tool
# - fullstack: Next.js + tRPC
# - microservice: Docker + K8s
# - ml-pipeline: MLOps setup
# - mobile-app: React Native
Custom Template Structure
templates/
├── node-api/
│ ├── template.yml
│ ├── src/
│ │ ├── index.ts
│ │ ├── routes/
│ │ └── middleware/
│ ├── tests/
│ ├── .env.example
│ └── README.md
MCP Configs (14)
Model Context Protocol server configurations:
Claude Context (Semantic Search)
{
"mcpServers": {
"claude-context": {
"command": "npx",
"args": ["-y", "@zilliz/claude-context"],
"env": {
"MILVUS_URI": "${MILVUS_URI}",
"OPENAI_API_KEY": "${OPENAI_API_KEY}"
}
}
}
}
Claude Code as MCP Server
{
"mcpServers": {
"claude-code": {
"command": "npx",
"args": ["claude-code-mcp"],
"env": {
"ANTHROPIC_API_KEY": "${ANTHROPIC_API_KEY}"
}
}
}
}
Custom MCP Configuration
{
"mcpServers": {
"custom-tools": {
"command": "node",
"args": ["./mcp-servers/custom/index.js"],
"env": {
"DATABASE_URL": "${DATABASE_URL}",
"REDIS_URL": "${REDIS_URL}"
}
}
}
}
Common Workflows
Complete Setup Flow
# 1. Install toolkit
/plugin marketplace add rohitg00/awesome-claude-code-toolkit
# 2. Install essential plugins
/plugin marketplace add rohitg00/pro-workflow
/plugin marketplace add buildingopen/bouncer
# 3. Set up security hooks
git clone https://github.com/emaarco/agento-patronum.git ~/.claude/hooks/security
# 4. Add cost optimization
npx skills add Sagargupta16/claude-cost-optimizer
# 5. Configure MCP servers
# Edit ~/.claude/config.json with MCP settings
# 6. Initialize project scaffolding
npx claude-scaffold init
Cost Optimization Workflow
# 1. Analyze current usage
ccusage analyze --last 30d
# 2. Enable cost-mode skill
/skill activate claude-cost-optimizer
# 3. Set up budget hooks
cat > ~/.claude/hooks/pre-approve.js << 'EOF'
module.exports = async (context) => {
const budget = 10.0; // USD per day
const spent = await context.usage.dailySpend();
if (spent > budget) {
throw new Error(`Daily budget exceeded: $${spent}`);
}
};
EOF
# 4. Run diagnostics
ccusage diagnose --top 10
Code Quality Workflow
# 1. Install quality gates
/plugin marketplace add buildingopen/bouncer
/plugin marketplace add hyhmrright/brooks-lint
# 2. Configure automatic audits
/skill activate bouncer-quick-audit
# 3. Run pre-commit hook
cat > .git/hooks/pre-commit << 'EOF'
#!/bin/bash
npx bouncer audit --quick
npx brooks-lint check --staged
EOF
chmod +x .git/hooks/pre-commit
Multi-Agent Project Management
# 1. Install project manager
/plugin marketplace add automazeio/ccpm
# 2. Create epic from issues
ccpm epic-start "Feature: User Authentication"
# 3. Activate agents
/agent activate backend-architect
/agent activate security-officer
/agent activate qa-engineer
# 4. Parallel execution
ccpm issue-analyze --parallel
# 5. Merge results
ccpm epic-merge
Configuration Examples
Settings.json Integration
{
"claude": {
"plugins": {
"autoload": [
"awesome-claude-code-toolkit",
"pro-workflow",
"agento-patronum"
]
},
"skills": {
"autoActivate": true,
"costMode": "balanced"
},
"hooks": {
"enabled": true,
"timeout": 5000
},
"agents": {
"maxConcurrent": 3,
"defaultModel": "claude-3-5-sonnet"
}
}
}
Memory Configuration
// memory/config.js
module.exports = {
storage: {
type: 'sqlite',
path: '~/.claude/memory/sessions.db'
},
compression: {
enabled: true,
model: 'claude-3-haiku',
targetRatio: 0.3
},
search: {
algorithm: 'hybrid', // BM25 + vector
topK: 5,
minScore: 0.7
},
retention: {
maxAge: '90d',
maxSize: '1GB'
}
};
Advanced Patterns
Custom Agent System
// agents/custom-agent.js
class CustomAgent {
constructor(config) {
this.name = config.name;
this.role = config.role;
this.tools = config.tools;
}
async execute(task) {
// Pre-task validation
await this.validate(task);
// Execute with error handling
try {
const result = await this.performTask(task);
await this.postProcess(result);
return result;
} catch (error) {
return this.handleError(error, task);
}
}
async validate(task) {
// Check prerequisites
if (!task.requirements) {
throw new Error('Task requirements missing');
}
// Verify permissions
const hasPermission = await this.checkPermissions(task);
if (!hasPermission) {
throw new Error('Insufficient permissions');
}
}
}
module.exports = CustomAgent;
Hook Chain Pattern
// hooks/chain.js
const hooks = [
require('./validate-security'),
require('./check-budget'),
require('./analyze-impact'),
require('./log-metrics')
];
module.exports = async (context) => {
for (const hook of hooks) {
await hook(context);
}
};
Skill Composition
---
name: fullstack-deploy
description: Complete fullstack deployment workflow
requires:
- backend-architect
- frontend-builder
- devops-engineer
---
# Fullstack Deploy
Orchestrates multiple skills for complete deployment:
1. **Backend**: Build and test API
2. **Frontend**: Build and optimize assets
3. **DevOps**: Deploy infrastructure
4. **QA**: Run smoke tests
5. **Monitor**: Set up observability
Troubleshooting
Plugin Not Loading
# Check plugin directory
ls -la ~/.claude/plugins/
# Verify plugin.json exists
cat ~/.claude/plugins/claude-code-toolkit/plugin.json
# Reload plugins
/plugin reload
# Check logs
tail -f ~/.claude/logs/plugins.log
Hook Execution Failures
# Test hook manually
node ~/.claude/hooks/pre-approve.js
# Enable debug mode
export CLAUDE_DEBUG=hooks
claude code
# Check hook permissions
chmod +x ~/.claude/hooks/*.js
Skill Not Activating
# List installed skills
/skill list
# Check skill triggers
cat ~/.claude/skills/*/SKILL.md | grep "triggers:"
# Force activation
/skill activate <name> --force
# Clear skill cache
rm -rf ~/.claude/cache/skills/
MCP Server Issues
# Test MCP server directly
npx @zilliz/claude-context --test
# Check environment variables
echo $MILVUS_URI
echo $OPENAI_API_KEY
# View MCP logs
tail -f ~/.claude/logs/mcp.log
# Restart MCP servers
/mcp restart
Cost Tracking Issues
# Verify JSONL files exist
ls -la ~/.claude/projects/*.jsonl
# Run diagnostic
ccusage diagnose --verbose
# Check for corrupted entries
ccusage validate --fix
# Export for manual analysis
ccusage export --format csv > usage.csv
Best Practices
1. Layered Configuration
~/.claude/
├── config.json # Global settings
├── plugins/ # Installed plugins
├── hooks/ # Lifecycle hooks
│ ├── global/ # Apply to all projects
│ └── project/ # Project-specific
├── skills/ # Installed skills
├── agents/ # Agent definitions
└── memory/ # Session memory
2. Cost Management
- Use Haiku for routine tasks (80% of work)
- Reserve Sonnet for complex logic
- Use Opus only for critical architecture decisions
- Enable prompt caching for repeated contexts
- Monitor with
ccusagedaily
3. Security Hardening
- Always use
agento-patronumhooks - Never commit
.envfiles - Use environment variable references
- Enable audit logging
- Regular security reviews with
bouncer
4. Quality Gates
- Pre-commit: Lint + format check
- Pre-approve: Security + cost validation
- Post-task: Quality audit with Bouncer
- Pre-merge: Brooks-lint review
5. Memory Management
- Compress sessions with AI summarization
- Regular cleanup of old sessions
- Index for fast retrieval
- Tag by project/feature for context
Resources
- Homepage: https://skillkit.sh
- Repository: https://github.com/rohitg00/awesome-claude-code-toolkit
- Documentation: https://github.com/rohitg00/awesome-claude-code-toolkit#readme
- Issues: https://github.com/rohitg00/awesome-claude-code-toolkit/issues
- License: Apache-2.0
- Stars: 1,700+ (16 stars/day)
Related Tools
- SkillKit: 400K+ skills marketplace
- Claude Mem: Automatic session memory
- CCPM: GitHub Issues project management
- Chief: Autonomous loop runner
- AgentLint: Repo AI-readiness checker
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/reason-machines/claude-code-skills/awesome-claude-code-toolkit">View awesome-claude-code-toolkit on skillZs</a>