Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
By mjohnson518
Orchestrate autonomous development workflows in Claude Code using coordinated agents and skills: design APIs/architectures, generate/review/refactor/debug/test code, audit security/performance/production readiness, optimize bottlenecks, manage git commits/branches/sprints across repos.
npx claudepluginhub mjohnson518/claude_superpowersSystem design and architecture mode - thorough, design-first approach
Teaching and learning mode - educational, explanatory approach
Fast development mode - speed-focused, get it working first
Code review mode - strict quality evaluation and feedback
This document describes how the 9 agents in claude-superpowers work together to handle complex development tasks.
Review code quality and best practices. Triggers after code changes, before commits, on PR reviews.
Systematic bug investigation and root cause analysis. Triggers on errors, test failures, crashes, unexpected behavior.
Generate technical documentation. Triggers on README requests, API docs, guides, architecture diagrams.
Execute git operations with user's configured identity. Triggers on commit requests, branch operations, git workflows.
REST, GraphQL, and API design best practices. Use when designing APIs, defining contracts, or reviewing API architecture.
Architectural patterns and design principles for scalable, maintainable systems. Use when designing systems, refactoring architecture, or choosing patterns.
Claude Code productivity patterns including EPCC workflow, context management, and extended thinking triggers. Use when starting sessions or managing context.
Performance profiling, optimization techniques, and bottleneck identification. Use when addressing performance issues or optimizing systems.
Enterprise-grade production readiness assessment system for comprehensive codebase evaluation. Use when (1) Evaluating a GitHub repository for production deployment, (2) Conducting pre-launch security and architecture reviews, (3) Assessing technical debt and system reliability, (4) Identifying gaps, vulnerabilities, and incomplete features, (5) Generating actionable remediation plans for engineering teams, (6) Validating scalability, observability, and operational readiness, (7) Reviewing cost optimization and resource efficiency, (8) Auditing compliance with industry standards (SOC2, GDPR, HIPAA, PCI-DSS), (9) Evaluating API contracts and integration stability, (10) Assessing team knowledge transfer and documentation completeness. Performs CTO-level multi-dimensional analysis exceeding top-tier tech company standards.
Uses power tools
Uses Bash, Write, or Edit tools
Share bugs, ideas, or general feedback.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Automated code review, security scanning, and quality enforcement
Agents for code review, security audits, debugging, and quality assurance
The most comprehensive Claude Code plugin — 14+ agents, 56+ skills, 33+ commands, and production-ready hooks for TDD, security scanning, code review, and continuous learning
The Paranoid Android — quality tools for code review, knowledge compounding, and work execution
Production safety hooks for autonomous Claude Code operation. Context monitoring, syntax checking, branch protection, activity logging, and more.
Harness engineering for Claude Code — hook-enforced dual review, state-machine gates, and fail-closed safety where it counts.
Production-ready tools for autonomous development with Claude Code.
| Component | Count | Description |
|---|---|---|
| Agents | 9 | Specialized AI agents for different tasks |
| Skills | 7 | Methodology guides and knowledge domains |
| Commands | 4 | Output style modes |
| Hooks | 2 | Automation triggers |
| Workflows | 1 | Ralph Loop autonomous iteration |
# Clone the repo
git clone https://github.com/mjohnson518/claude_superpowers.git
# Install as a plugin (coming soon)
claude plugins install ./claude_superpowers
# Clone the repo
git clone https://github.com/mjohnson518/claude_superpowers.git
# Copy to your project's .claude directory
cp -r claude_superpowers/skills/* /path/to/your/project/.claude/skills/
cp -r claude_superpowers/agents/* /path/to/your/project/.claude/agents/
cp -r claude_superpowers/commands/* /path/to/your/project/.claude/commands/
cp -r claude_superpowers/hooks/* /path/to/your/project/.claude/hooks/
Nine specialized agents for different development tasks:
| Agent | Purpose | Trigger Keywords |
|---|---|---|
| orchestrator | Coordinate complex multi-step tasks | "improve", "refactor", large features |
| code-reviewer | Quality assessment (0-10 scoring) | "review", "check quality", before commit |
| security-auditor | OWASP-based vulnerability detection | "security scan", "audit", auth changes |
| debugger | Root cause analysis | "bug", "error", "not working" |
| docs-writer | Documentation generation | "document", "README", "API docs" |
| test-runner | Test/lint/build validation | "run tests", "validate", "CI" |
| git-executor | Git operations | "commit", "branch", "push" |
| refactorer | Code structure improvement | "clean up", "technical debt" |
| test-architect | Test strategy design | "test strategy", "coverage" |
The git-executor agent enforces commit standards. Configure your identity:
# Set your git identity (required before first use)
git config user.name "Your Name"
git config user.email "your-email@example.com"
See agents/ORCHESTRATION.md for workflow diagrams.
Seven methodology guides and knowledge domains:
| Skill | Purpose |
|---|---|
| production-readiness | 15-dimension production audit with scoring |
| claude-code-optimizer | EPCC workflow (Explore → Plan → Code → Commit) |
| sprint-orchestrator | Multi-project sprint management |
| Skill | Purpose |
|---|---|
| architecture-patterns | Clean Architecture, DDD, microservices |
| performance-optimization | Profiling, caching, bottleneck identification |
| api-design | REST, GraphQL, and API best practices |
| project-analysis | Codebase analysis and onboarding |
Four output style modes to adjust Claude's behavior:
| Command | Mode | Use Case |
|---|---|---|
/superpowers:architect | System design | Architecture decisions, design docs |
/superpowers:rapid | Fast development | Prototyping, getting things working |
/superpowers:mentor | Teaching | Learning, explanations, tutorials |
/superpowers:review | Code review | Quality assessment, PR reviews |
Automation triggers for security and file protection:
.env, .git/, lock files, *.pem, *.keyHooks are configured in hooks/hooks.json. To disable:
{
"hooks": [
{
"name": "security-scan",
"enabled": false
}
]
}
Logs are written to .claude/logs/.
Autonomous iterative development loop - a working alternative to the broken ralph-wiggum plugin.
Features:
Quick Start:
# Copy files to your project
cp workflows/ralph-loop/ralph-loop.sh /path/to/your/project/
cp workflows/ralph-loop/PROMPT_TEMPLATE.md /path/to/your/project/.claude/RALPH_PROMPT.md
# Edit the prompt
nano .claude/RALPH_PROMPT.md
# Run (50 iterations max)
./ralph-loop.sh 50
See workflows/ralph-loop/README.md for complete documentation.