Development framework for Claude Code with specialized agents and automated workflows.

📚 View Full Documentation

uvx --from git+https://github.com/rysweet/MicrosoftHackathon2025-AgenticCoding amplihack launch

Launches Claude Code with preconfigured agents. No installation needed.

• Quick Start
  • Prerequisites
  • Basic Usage
  • Create Alias for Easy Access
• Core Concepts
  • Workflow
  • Philosophy
• Configuration
  • Anthropic (Default)
  • Azure OpenAI
  • GitHub Copilot CLI
  • Custom Workflows
• Commands Reference
• Agents Reference
  • Core Agents
  • Specialized Agents
• Features
  • Workflow Orchestration by Default
  • Goal-Seeking Agent Generator
  • Profile Management
  • GitHub Pages Documentation Generation
  • Additional Features
  • Statusline
• Documentation
  • Getting Started
  • Features
  • Patterns
  • Configuration
  • Development
  • Methodology
  • Security
  • Core Principles
• Development
  • Contributing
  • Local Development
  • Testing
• License

• Python 3.8+, Node.js 18+, npm, git
• GitHub CLI (gh) for PR/issue management
• uv (astral.sh/uv)

For detailed installation instructions, see docs/PREREQUISITES.md.

# Launch Claude Code with amplihack
amplihack launch

# With Azure OpenAI (requires azure.env configuration)
amplihack launch --with-proxy-config ./azure.env

# Work directly in a GitHub repository
amplihack launch --checkout-repo owner/repo

Not sure where to start? Use the command above to run from uvx, then tell Claude Code to cd /path/to/my/project and provide your prompt. All prompts are automatically wrapped with /amplihack:ultrathink for workflow orchestration (use --no-ultrathink flag to opt-out for simple tasks).

Instead of typing the full uvx command, create an alias:

# Add to your ~/.bashrc or ~/.zshrc
alias amplihack='uvx --from git+https://github.com/rysweet/MicrosoftHackathon2025-AgenticCoding amplihack'

# Reload your shell
source ~/.bashrc  # or source ~/.zshrc

Now you can simply run:

amplihack launch
amplihack launch --with-proxy-config ./azure.env
amplihack launch --checkout-repo owner/repo

Iterative multi-step development process (customizeable via DEFAULT_WORKFLOW.md)

1. Clarify requirements
2. Create issue
3. Setup branch
4. Design tests
5. Implement
6. Simplify
7. Test
8. Commit
9. Create PR
10. Review
11. Integrate feedback
12. Check philosophy
13. Prepare merge
14. Cleanup

• Simplicity - Start simple, add only justified complexity
• Modular - Self-contained modules with clear interfaces
• Working code - No stubs or dead code
• Test-driven - Tests before implementation

amplihack works with Claude Code and Anthropic models by default. For additional capabilities, you can configure Azure OpenAI integration.

amplihack works with Claude Code and Anthropic models out of the box. No additional configuration needed.

To use Azure OpenAI models, create an azure.env file with the following minimum configuration:

# Required: Your Azure OpenAI API key
AZURE_OPENAI_API_KEY=your-api-key

# Required: Azure OpenAI endpoint (base URL without path)
AZURE_OPENAI_ENDPOINT=https://your-resource.openai.azure.com

# Required: Model deployment name (use either BIG_MODEL or AZURE_OPENAI_DEPLOYMENT_NAME)
AZURE_OPENAI_DEPLOYMENT_NAME=gpt-4.1

# Optional: API version (defaults to 2025-04-01-preview)
AZURE_OPENAI_API_VERSION=2024-12-01-preview

Launch with Azure configuration:

amplihack launch --with-proxy-config ./azure.env

Note: The endpoint should be just the base URL (e.g., https://your-resource.openai.azure.com) without /openai or other path suffixes. The proxy will automatically construct the correct API paths.

Security Warning: Never commit API keys to version control. Use environment variables or secure key management systems.

amplihack also supports GitHub Copilot CLI integration. See docs/github-copilot-litellm-integration.md for setup instructions.

The iterative-step workflow is fully customizable. Edit .claude/workflow/DEFAULT_WORKFLOW.md to modify the development process - changes apply immediately to /ultrathink and other commands. See docs/WORKFLOW_COMPLETION.md for detailed customization instructions.

Distribute agentic work across Azure VMs:

amplihack remote auto "implement feature" --region westus3 --vm-size s

Documentation: .claude/tools/amplihack/remote/README.md

All prompts are automatically wrapped with /amplihack:ultrathink for maximum effectiveness. This enables:

• Multi-agent workflow orchestration
• Multi-step development workflow
• Automated architecture, building, and testing
• Philosophy compliance checking

Benchmark results: Amplihack without orchestration = vanilla Claude. The orchestration IS the value! See benchmarking guide for measuring performance.

Opt-out for simple tasks:

# Skip orchestration with --no-ultrathink flag
amplihack launch --no-ultrathink -- -p "simple prompt"

# Or use slash commands directly
amplihack launch -- -p "/analyze src/file.py"

How it works:

# Before: Manual orchestration required
amplihack launch -- -p "/amplihack:ultrathink implement feature"

# Now: Automatic orchestration (same result)
amplihack launch -- -p "implement feature"

Create autonomous agents from simple prompts:

# Write your goal
cat > my_goal.md <<'EOF'
# Goal: Automated Code Review
Review Python code and suggest improvements.
EOF

# Generate agent
amplihack new --file my_goal.md

# Run agent
cd goal_agents/automated-code-review-agent
python main.py

Features:

• Generate agents in < 0.1 seconds
• Automatic skill matching
• Multi-phase execution planning
• Standalone, distributable agents

Learn more: Goal Agent Generator Guide

Reduce token usage by 72% with profile-based component filtering:

# Install with filtering
amplihack install

# Result: Only 9/32 agents staged (72% reduction)

# Launch with filtering
amplihack launch

# Result: Focused environment for coding tasks

Built-in Profiles:

• all: Full environment (32 agents, default)
• coding: Development-focused (9 agents)
• research: Investigation-focused (7 agents)

Learn more: Profile Management Guide

Generate professional documentation sites automatically:

• Auto-discovers content from docs/, README.md, and .claude/commands/
• Three-pass validation ensures quality documentation
• Safe gh-pages deployment with rollback support
• Local preview server for testing
• MkDocs + Material theme integration

Learn more:

• Tutorial: Your First Documentation Site
• How-To: Generate GitHub Pages Sites
• API Reference: GitHub Pages Module

• Power-Steering - AI-powered session guidance with intelligent redirect detection (🚦 indicator)
• Auto Mode - Autonomous agentic loops for multi-turn workflows (/amplihack:auto)
• Lock Mode - Continuous work mode without stopping (/amplihack:lock, /amplihack:unlock) (🔒 indicator)
• Document-Driven Development - Systematic methodology for large features with documentation-first approach
• Fault-Tolerant Workflows - N-version programming, multi-agent debate, and cascade fallback patterns
• Security Analysis - XPIA cross-prompt injection defense (/amplihack:xpia)
• Neo4j Memory System - Persistent memory and knowledge graphs across sessions
• Investigation Workflow - Deep knowledge excavation with historical context
• Skills System - 54+ skills including PDF, XLSX, DOCX, PPTX, analysts, and workflow patterns
• Fix Workflow - Rapid resolution of common error patterns (/amplihack:fix)
• Reflection & Improvement - Session analysis and learning capture (/amplihack:reflect, /amplihack:improve)
• Socratic Questioning - Challenge claims and clarify requirements (/amplihack:socratic)
• Expert Panel - Multi-expert review with voting (/amplihack:expert-panel)
• Knowledge Builder - Build comprehensive knowledge base (/amplihack:knowledge-builder)
• Transcripts Management - Conversation transcript tracking (/amplihack:transcripts)
• Modular Build - Self-contained modules with clear contracts (/amplihack:modular-build)
• Pre-commit Diagnostics - Fix formatting, linting, type checking before push
• CI Diagnostics - Monitor CI, diagnose failures, iterate until mergeable
• Worktree Management - Git worktree automation for parallel development
• Session Logs - Comprehensive logging and decision records
• Customization System - Manage user preferences (/amplihack:customize)

Real-time session information displayed at the bottom of Claude Code showing:

• Current directory and git status (branch, clean/dirty)
• Active model (Opus/Sonnet/Haiku)
• Token usage 🎫, Cost 💰, and Duration ⏱
• Feature indicators: Power-Steering 🚦, Lock Mode 🔒

Example:

~/src/amplihack (main → origin) Sonnet 🎫 234K 💰$1.23 ⏱12m

Full documentation: docs/reference/STATUSLINE.md

• Prerequisites - Platform setup and dependencies
• Proxy Configuration - Azure OpenAI proxy setup

• Auto Mode - Autonomous agentic loop
• Agent Bundles - Custom agent creation
• GitHub Copilot Integration - Copilot CLI support
• Office Skills - PDF, Excel, Word, and PowerPoint document processing
  • PDF Skill - Comprehensive PDF manipulation
  • XLSX Skill - Spreadsheet creation with formulas and financial modeling
• Azure Admin Skill - Azure administration, identity management, RBAC, and resource orchestration
• Azure DevOps CLI Skill - Azure DevOps automation, pipelines, repos, and artifacts
• Azure DevOps Boards Tools - Work item management with CLI tools (Quick Start)
• Benchmarking with eval-recipes - Performance measurement and comparison
• Profile Management - Token optimization and environment customization

• Workspace Pattern - Multi-project organization with git submodules

• Hook Configuration - Session hooks
• Workflow Customization - Process customization

• Developing amplihack - Contributing guide
• Implementation Summary - Architecture overview

• Document-Driven Development - Systematic approach for large features
• DDD Overview - Comprehensive guide to DDD principles
• Core Concepts - Context poisoning, file crawling, retcon writing
• DDD Phases - Step-by-step implementation guide

• Security Recommendations - Best practices
• Security Context Preservation - Context handling

• The Amplihack Way - Effective strategies for AI-agent development
• Discoveries - Documented problems, solutions, and learnings
• Creating Tools - Build custom AI-powered tools
• Philosophy - Core principles and patterns
• Workflows - Development process

Fork, submit PRs. Add agents to .claude/agents/, patterns to .claude/context/PATTERNS.md.

git clone https://github.com/rysweet/MicrosoftHackathon2025-AgenticCoding.git
cd MicrosoftHackathon2025-AgenticCoding
uv pip install -e .
amplihack launch

pytest tests/

Amplihack supports RustyClawd, a high-performance Rust implementation of Claude Code.

# Force RustyClawd usage explicitly
amplihack RustyClawd -- -p "your prompt"

# Or set environment variable
export AMPLIHACK_USE_RUSTYCLAWD=1
amplihack launch -- -p "your prompt"

• 5-10x faster startup compared to Node.js Claude Code
• 7x less memory usage
• Rust safety guarantees - no runtime errors
• Same features - drop-in compatible

RustyClawd must be available in your system PATH:

Option 1: Install via cargo

cargo install --git https://github.com/rysweet/RustyClawd rusty

Option 2: Build from source

git clone https://github.com/rysweet/RustyClawd
cd RustyClawd
cargo build --release
# Add to PATH or use RUSTYCLAWD_PATH environment variable
export RUSTYCLAWD_PATH=$PWD/target/release/rusty

Option 3: Custom binary location

# Point to your custom RustyClawd build
export RUSTYCLAWD_PATH=/path/to/your/rusty

• AMPLIHACK_USE_RUSTYCLAWD: Force RustyClawd usage (1/true/yes)
• RUSTYCLAWD_PATH: Custom path to RustyClawd binary (optional)

MIT. See LICENSE.