DOCSEARCHbyAlgolia
Submit Docs

SuperClaude Framework

repository·master·Indexed 12 days ago

https://github.com/superclaude-org/superclaude_framework

A context-oriented configuration framework that extends Claude Code with structured development workflows, specialized AI agents, and MCP server integrations. It utilizes Markdown instruction files to modify AI behavior via slash commands (/sc:[command]) and agent handles (@agent-[name]), and includes a memory system for error learning and pattern capture.

Tokens
269.9K
Snippets
678
Records
1.1K
Agent score
95%

What's inside SuperClaude

  1. Overview of available MCP Servers

    master

    Model Context Protocol (MCP) servers provide enhanced capabilities to the SuperClaude framework.

    Core MCP Servers

    • Context7: Official documentation lookup and framework patterns.
    • Sequential: Multi-step reasoning and complex analysis.
    • Magic: Modern UI component generation from 21st.dev patterns.
    • Playwright: Browser automation and E2E testing.
    • Morphllm: Pattern-based code editing with token optimization.
    • Serena: Semantic code understanding and project memory.

    System Requirements

    • Node.js: version 16.0.0 or higher.
    • Package Manager: npm or yarn.
    • Network: Stable connection (required for some servers).
    • Memory: 2GB+ recommended.

  2. Overview of SuperClaude Behavior Modes

    master

    SuperClaude uses different behavior modes to adapt its reasoning, communication style, and tool usage based on task complexity. Modes can be automatically activated by specific keywords or task characteristics, or manually triggered using flags or commands.

    Mode Summary Table

    ModePurposeAuto-Trigger ExamplesKey BehaviorBest For
    🧠 BrainstormInteractive discovery"maybe", "brainstorm", vague requestsSocratic questioning, requirement elicitationNew project planning, unclear requirements
    🔍 IntrospectMetacognitive analysisError recovery, "analyze reasoning"Transparent thought markers (🤔, 🎯, 💡)Debugging, learning, optimization
    📋 Task ManagementComplex coordination>3 steps, >2 directoriesPhase decomposition, session persistenceMulti-step operations, project management
    🎯 OrchestrationIntelligent tool selectionMulti-tool operations, high resource useOptimal tool routing, parallel executionComplex analysis, performance optimization
    ⚡ Token EfficiencyCompressed communicationHigh context usage, --uc flagSymbol systems, 30-50% token reductionResource constraints, large-scale operations
    🎨 StandardBalanced defaultSimple, well-defined tasksProfessional, clear technical languageGeneral development tasks

    Quick Verification

    You can test modes using the /sc: command prefix (e.g., /sc:brainstorm "mobile app").

    # Automatic activation examples
/sc:brainstorm "mobile app"        # → Socratic discovery
/sc:implement "authentication system" # → Multi-phase coordination
"--uc analyze large-codebase/"     # → Compressed symbol output

  3. Overview of SuperClaude Framework Capabilities

    master

    SuperClaude is a metaprogramming configuration framework that transforms Claude Code into a structured development platform through instruction injection and component control.

    Framework Statistics:

    • 30 Slash Commands: Covering the full development lifecycle from brainstorming to deployment.
    • 16 Specialized Agents: Domain-specific AI experts (e.g., PM Agent, Security Engineer, Frontend Architect).
    • 7 Operating Modes: Adaptive contexts including Brainstorming, Business Panel, Deep Research, Orchestration, Token Efficiency, Task Management, and Reflection.
    • 8 MCP Server Integrations: For enhanced capabilities like web search, memory, and UI component generation.

  4. Overview of SuperClaude Command Categories

    master

    The SuperClaude Framework provides 30 slash commands categorized by their role in the development lifecycle. Understanding these categories helps you choose the right tool for your current goal:

    1. ORCHESTRATION (Meta-level): High-level planning and task management (e.g., /sc:pm, /sc:spawn, /sc:task, /sc:workflow).
    2. DISCOVERY: Requirements gathering and research (e.g., /sc:brainstorm, /sc:research).
    3. IMPLEMENTATION: Building features and designing systems (e.g., /sc:implement, /sc:design).
    4. QUALITY: Verification and diagnosis (e.g., /sc:analyze, /sc:troubleshoot, /sc:test, /sc:build).
    5. IMPROVEMENT: Refactoring and maintenance (e.g., /sc:improve, /sc:cleanup).
    6. DOCUMENTATION: Explaining and indexing code (e.g., /sc:explain, /sc:document, /sc:index-repo).
    7. EXPERT PANELS: Specialized reviews (e.g., /sc:spec-panel, /sc:business-panel).
    8. UTILITIES: General development tasks (e.g., /sc:git, /sc:estimate, /sc:reflect, /sc:load, /sc:save).

  5. Overview of SuperClaude Behavioral Modes

    master

    SuperClaude uses different behavioral modes to adapt Claude's reasoning and communication style to your specific task. Modes can be activated automatically by task patterns or manually via flags.

    Mode Reference Table

    ModePurposeAuto-TriggersKey BehaviorsBest Used For
    🧠 BrainstormingInteractive discovery"brainstorm", "maybe", vague requestsSocratic questions, requirement elicitationNew project planning, unclear requirements
    🔍 IntrospectionMeta-cognitive analysisError recovery, "analyze reasoning"Transparent thinking markers (🤔, 🎯, 💡)Debugging, learning, optimization
    🔬 Deep ResearchSystematic investigation mindset/sc:research, investigation keywords6-phase workflow, evidence-based reasoningTechnical research, current events, market analysis
    📋 Task ManagementComplex coordination>3 steps, >2 directoriesPhase breakdown, memory persistenceMulti-step operations, project management
    🎯 OrchestrationIntelligent tool selectionMulti-tool ops, high resource usageOptimal tool routing, parallel executionComplex analysis, performance optimization
    ⚡ Token EfficiencyCompressed communicationHigh context usage, --uc flagSymbol systems, estimated 30-50% token reductionResource constraints, large operations

  6. Developer Guide Index for SuperClaude Framework

    master

    The SuperClaude Framework developer guide is organized into three primary tracks based on your role: contributing new context files, designing system architecture, or verifying installations and troubleshooting issues.

    Contributing Context Files

    If you are developing new context files or slash commands, follow this workflow:

    1. Setup: Configure your development environment using the Development Setup.
    2. Development: Adhere to the Context File Guidelines when creating new files.
    3. Validation: Ensure files meet framework standards using File Validation.
    4. Support: Refer to Getting Help for assistance.

    Architecting Systems

    If you are designing how agents and commands interact within the framework, use these resources:

    1. Architecture: Understand the Context File Architecture.
    2. Patterns: Study the Agent and Command Structure for component design.
    3. Extensibility: Learn how to use Extending the Framework.
    4. Integration: Configure Model Context Protocol (MCP) servers via MCP Configuration.

    Verification and Troubleshooting

    If you are verifying an installation or debugging issues:

    1. Installation: Run Installation Verification.
    2. Validation: Perform Context File Verification.
    3. Troubleshooting: Consult Common Issues and use Troubleshooting Commands for diagnostics.

  7. Understand the SuperClaude development phases

    master

    The SuperClaude framework development is organized into several distinct phases aimed at transforming it into a self-improving platform:

    1. Phase 1: Documentation Structure (Completed): Established the docs/Development/ directory and core architecture documentation.
    2. Phase 2: PM Agent Mode Integration (In Progress): Integrating the PM Agent as an active meta-layer using the PDCA cycle and session lifecycle management.
    3. Phase 3: Serena MCP Integration (Planned): Implementing full session persistence and memory operations via Serena MCP.
    4. Phase 4: Documentation Strategy (Planned): Implementing a systematic documentation lifecycle (temp/patterns/mistakes) and automated knowledge management.
    5. Phase 5: Auto-Activation System (Research): Researching ways to have the PM Agent activate automatically at every session start with minimal startup delay (<500ms).

  8. Use the Serena MCP Server for semantic code understanding

    master

    The Serena MCP Server is designed for semantic code understanding, providing project memory and session persistence. It is the preferred tool for operations involving symbol manipulation, dependency tracking, and large-scale architectural analysis.

    Key Capabilities

    • Symbol Operations: Renaming, extracting, or moving functions and classes.
    • Code Navigation: Project-wide exploration and finding symbol references.
    • LSP Integration: Handling multi-language projects through Language Server Protocol integration.
    • Session Management: Using /sc:load to activate project context and /sc:save to persist memory and work sessions.
    • Large Codebase Analysis: Managing complex architectures and codebases exceeding 50 files.

  10. Explore the SuperClaude Documentation structure

    master

    The SuperClaude documentation is organized into four main categories to help users navigate the framework:

  11. SuperClaude Slash Commands Overview

    master

    SuperClaude provides a suite of specialized slash commands designed to extend Claude Code's capabilities. While some commands overlap with standard Claude Code functionality, SuperClaude's versions are often enhanced with multi-agent coordination, MCP (Model Context Protocol) integration, and specific persona-driven workflows.

    Key Command Categories

    • High-Value Unique Commands (❌): Commands that provide unique value not found in standard Claude Code, such as brainstorm (multi-agent coordination), design (architecture protocols), pm (PM Agent core), reflect (reflexion loops), and spawn (sub-agent branching).
    • Enhanced/Specialized Commands (⚠️): Commands that share names with standard Claude Code but are significantly enhanced for specific workflows:
      • build: DevOps-focused with Playwright MCP integration and report generation.
      • cleanup: Includes multi-persona checks (Architect/Quality/Security) and safe rollbacks.
      • explain: Educational focus with persona-driven level control.
      • implement: Large-scale flows using multiple MCPs (Context7, Magic, Playwright) and multi-persona generation.
      • improve: Specialized improvement loops for quality, performance, maintainability, and security.
      • research: Advanced multi-hop research using Tavily, Serena, and Playwright MCPs.
      • task: Advanced PM-centric task management (Epic → Story → Task).
      • test: QA-focused testing with Playwright MCP and automatic repair suggestions.
    • Redundant Commands (✅): Commands like git that are functionally identical to standard Claude Code and are candidates for removal or integration.

  12. Explore SuperClaude Recipe Collections

    master

    The SuperClaude Examples Cookbook is organized into three distinct collections based on user expertise and project complexity:

    1. Basic Examples Collection: Focuses on essential commands, single-agent workflows, and copy-paste ready solutions for common development tasks. Best for new users.
    2. Advanced Workflows Collection: Focuses on multi-agent coordination, enterprise-scale project workflows, and session management. Best for experienced users.
    3. Integration Patterns Collection: Focuses on framework-specific integration, performance optimization, and cross-tool coordination. Best for system architects and experts.