Skip to main content
Acolytes for Claude Code Logo

Revolutionary Multi-Agent System v2.0.0 with

♾️ NEURAL MEMORY NETWORK

Transform Claude Code into an intelligent project orchestrator with 60 specialized agents + 5 setup agents + project-specific Acolytes. Features revolutionary Acolytes Quest System for unprecedented multi-agent coordination and persistent SQLite memory for cumulative learning.

PyPI version Agents Quest Memory Hooks Status

🌍 FIRST IN THE WORLD TO ACHIEVE THIS

🚀 BREAKTHROUGH: Acolytes Quest System

NEVER BEFORE ACHIEVED: Acolytes for Claude Code features the world's first stateful multi-agent coordination system that enables real-time communication and collaboration between 55 specialized agents + 5 setup agents. The Quest system operates through persistent SQLite database, ensuring no information loss and perfect autonomous coordination.

Quest System Success Stories

  • ✅ Real-time Agent Communication: Agents communicate through persistent SQLite database
  • ✅ Autonomous Coordination: Turn-based system prevents conflicts and ensures perfect handoffs
  • ✅ Stateful Memory: Project Acolytes maintain 14 types of persistent memory
  • ✅ Enterprise Reliability: Production-tested coordination with complete audit trails

Quest System Workflow Example (Fully Automated)

USER> "/prequest implement OAuth authentication"
  ↓
Claude → 🎭 @coordinator.backend creates Quest #1847
  ↓
📞 @service.auth: "Building OAuth flow... need users schema"
  ↓
📞 @database.postgres: "Creating oauth_provider, refresh_token columns"
  ↓
🎭 @coordinator.backend: "Schema ready, continue OAuth implementation"
  ↓
📞 @service.auth: "OAuth flow complete with Google & GitHub"
  ↓
Claude → 🎯 "Quest completed: OAuth system ready"

🤖 Everything happens automatically - no manual coordination needed!

🏗️ Advanced Multi-Agent Architecture

58 Global Agents

  • Backend: API, Go, Java, Laravel, Node.js, Python, Rust, Serverless
  • Frontend: Angular, React, Vue, Mobile
  • Database: PostgreSQL, MongoDB, Redis, MariaDB, SQLite, pgvector, PostGIS, Vectorial
  • DevOps: Bash, CI/CD, Containers, Git, IaC, Monitoring, Performance, Troubleshooting, Webserver
  • Business: Billing, Payments, Subscriptions
  • Services: AI/ML, Auth, Communication, Data Processing, Integrations, Mapbox
  • Coordinators: Backend, Frontend, Database, DevOps, Infrastructure, Migration, Security, Testing
  • Analysis: Data Analysis, Strategic Planning, Compliance, Security Audit
  • Documentation: Specialist
  • Testing: Quality Assurance
  • System: Setup (5 agents), Planning

Acolytes (Auto-Created per Project)

  • One expert agent per detected module (acolyte.auth, acolyte.api, etc.)
  • Deep module knowledge stored in 14-memory SQLite system
  • Created during /setup Phase 6 via @setup.acolytes-creator

🎯 Intelligent Agent Routing System

17-rule anti-ambiguity system eliminates agent selection errors:

  • Strategy vs Implementation: Coordinators for decisions, specialists for execution
  • RAG/Vector Search: PostgreSQL+embeddings → @database.pgvector, Any vector platform → @database.vectorial
  • Auth vs Security: OAuth/JWT → @service.auth, Architecture/compliance → @coordinator.security
  • Multi-Agent Workflows: Sequential, parallel, and coordinated execution patterns

📖 IMPORTANT: READ DOCUMENTATION FIRST

🚨 CRITICAL: Before using Acolytes, you MUST read the complete documentation:

These guides contain essential information for proper usage. Do not skip them.


⚡ Quick Start

Installation

Option 1: Install from PyPI (Recommended)

# Install Acolytes package
pip install acolytes

# Initialize Acolytes system (copies all files to ~/.claude)
acolytes --init

# Navigate to YOUR project
cd /path/to/your/project

# Open Claude with safety bypass (⚠️ hooks protect against malicious commands)
claude --dangerously-skip-permissions -c

# Inside Claude, run setup
/setup

Option 2: Install from Source

# Clone repository
git clone https://github.com/unmasSk/Acolytes.git
cd Acolytes

# Install package
pip install -e .

# Initialize Acolytes system
acolytes --init

# Navigate to YOUR project
cd /path/to/your/project

# Open Claude with safety bypass (⚠️ hooks protect against malicious commands)
claude --dangerously-skip-permissions -c

# Inside Claude, run setup
/setup

Available Commands

Command Description Agent Integration
/setup Integrates Acolytes into your project (new or existing) - 6-phase analysis 5 setup agents + Acolytes creation
/commit Smart commits with auto-versioning, changelog, and git tagging @docs.specialist + bump2version
/pr Create pull request with comprehensive analysis @ops.git + review agents
/acolytes Process agent coordination (IN DEVELOPMENT) Quest system coordinator
/save Save current session to SQLite database Session persistence

Claude auto-selects the right agent based on context:

  • USER> "Optimize my API endpoints" → Claude automatically invokes @backend.python or @backend.nodejs
  • USER> "The authentication is broken" → Claude knows to invoke @service.auth
  • USER> "Add payment processing to my app" → Claude orchestrates @business.payment + @database.postgres + @backend.api

Or you can explicitly request an agent:

  • USER> "Use @database.postgres to fix the indexing issues" → Claude directs task to specified agent
  • USER> "Ask @coordinator.security about compliance requirements" → Claude consults the security coordinator

🏗️ Revolutionary Architecture

Main Claude Session (Orchestrator)
    │
    ├── 🌍 GLOBAL AGENTS (58) - ~/.claude/agents/
    │   ├── 💻 Backend & Frontend Engineers
    │   ├── 🎛️ Coordinators: Strategic architects for complex decisions
    │   ├── 💾 Database Experts: All major databases + vector stores
    │   ├── ⚙️ DevOps & Operations: CI/CD, monitoring, performance
    │   ├── 💼 Business Systems: Payments, billing, subscriptions
    │   ├── 🔍 Analysis & Audit: Security, compliance, data analysis
    │   └── 🔧 System Agents: Setup, planning
    │
    ├── 🏠 PROJECT ACOLYTES - project/.claude/agents/
    │   ├── acolyte.auth (YOUR authentication module)
    │   ├── acolyte.api (YOUR API endpoints)
    │   └── [Auto-created for each detected module]
    │
    └── 🗄️ PERSISTENT MEMORY (SQLite + MCP)
        ├── Session Memory: Jobs system groups Claude sessions
        ├── Acolyte Memory: 14 types per project Acolyte
        └── Full audit trail of all operations

🗄️ Persistent Memory System

SQLite Database (8 Active Tables)

  1. agents_catalog - Directory of all 60 agents with routing rules
  2. agents_memory - Deep knowledge storage for each agent/module
  3. sessions - Conversation history with context preservation
  4. messages - Individual message tracking with metadata
  5. tool_logs - Execution history for optimization
  6. todos - Task management synced with Claude
  7. acolyte_quests - Quest coordination data
  8. acolyte_status_changes - Agent status and communication logs

In Development: agent_health - Agent monitoring and performance metrics

14 Agent Memory Types

  • knowledge - Core understanding: purpose, features, architecture, TODOs
  • structure - Code organization: files, classes, functions, APIs
  • patterns - Best practices: conventions, design patterns, anti-patterns
  • interfaces - Module exports: public APIs, events, contracts
  • dependencies - Connections: internal/external deps, services
  • schemas - Data models: entities, validation, transformations
  • quality - Code health: tests, coverage, security analysis
  • operations - DevOps: deployment, monitoring, CI/CD, migrations
  • context - Business logic: decisions, history, roadmap
  • domain - Specialized: ML models, GraphQL, i18n
  • security - Security profile: permissions, compliance, vulnerabilities
  • errors - Error handling: common failures, recovery procedures
  • performance - Optimization: bottlenecks, caching, scaling
  • history - Recent interactions: consultations, implementations

MCP Server Integration

Required MCPs:

mcp-code-index: # Fast code search and indexing
mcp-server-fetch: # External API interactions
sequential-thinking: # Complex reasoning chains
playwright: # Browser automation and testing
context7: # Advanced context management

Optional MCPs:

mcp-server-sqlite: # Database operations (if needed)
mcp-server-git: # Git operations (if needed)
n8n-mcp: # Workflow automation
21st-dev_magic: # UI/UX enhancements

🚀 Real-World Usage Examples

Example 1: Authentication System (Automatic Orchestration)

USER> "Implement OAuth authentication with Google and GitHub"

Claude automatically orchestrates:

  1. @coordinator.security → Recommends OAuth2 + JWT hybrid approach
  2. @service.auth → Implements OAuth flow with passport.js
  3. @database.postgres → Creates users, sessions, oauth_providers tables
  4. @backend.nodejs → Integrates middleware and endpoints
  5. @frontend.react → Builds login components with provider buttons

Result: Complete OAuth system in one request, all coordinated automatically

Example 2: Performance Optimization

Example 3: Complex Migration (Multi-Agent Coordination)

USER> "Migrate from MongoDB to PostgreSQL"

Coordinated execution:

  1. @coordinator.migration → Creates migration strategy
  2. @database.mongodb + @database.postgres → Schema mapping
  3. Data type mapping coordinated automatically
  4. @backend.nodejs → Updates all queries and models
  5. @ops.monitoring → Sets up migration monitoring
  6. Rollback plan stored in SQLite for safety

All decisions preserved in Jobs system for future reference

🗺️ Roadmap to Revolution

Current Version: v1.0.7 ✨ → Target Version: v2.0.0 🚀

🎯 8-Phase Evolution (10 Months)

🎭 Phase 1
v2.0.0
8 weeks

Quest System Revolution
Multi-agent coordination, stateful communication, turn-based orchestration
📊 Phase 2
v2.1.0
2 weeks

Dashboard & Control Center
Real-time agent monitoring, Quest visualization, performance metrics
📋 Phase 3
v2.2.0
3 weeks

Templates & Patterns
8 specialized templates: E-commerce, SaaS, API, Mobile, ML, CMS
📦 Phase 4
v2.3.0
2 weeks

Export/Import & Portability
Configuration portability, backup/restore, team sharing
🔄 Phase 5
v2.4.0
2 weeks

Auto-Update & Intelligence
Zero-downtime updates, smart rollouts, community agents
🧠 Phase 6
v2.5.0
3 weeks

Knowledge Management
Cross-project learning, pattern recognition, collective intelligence
🏢 Phase 7
v2.6.0
4 weeks

Enterprise Features
Code review automation, notifications, achievements, RBAC
🌟 Phase 8
v3.0.0
4 weeks

Universal AI Platform
Semantic routing, enhanced parallel spawning, universal MCP, multi-LLM support

🎯 What's Coming Next

🔥 Priority 1 - Dashboard Revolution (v2.1.0)

/acolytes --dashboard  # Real-time command center with live agent monitoring
  • 📈 Real-time metrics - Performance, success rates, agent health
  • 🎯 Quest visualization - Live coordination system monitoring
  • 🎯 Agent status grid - All 60 agents at a glance
  • ⚡ <500ms rendering - Lightning-fast dashboard experience

🚀 The V2.0.0 Vision:

  • Stateful multi-agent orchestration - First-ever persistent agent coordination
  • Zero context loss - Perfect memory continuity across sessions
  • Turn-based communication - Conflict-free agent collaboration
  • Autonomous task completion - Agents work together without human intervention

"From Revolutionary v1.0.7 to Legendary v2.0.0 - The Quest System Era"

🌟 Why Acolytes for Claude Code is Revolutionary

Enterprise-Grade Multi-Agent Coordination

No other system provides seamless cross-agent coordination through persistent SQLite database communication. This enables zero information loss during agent handoffs and perfect task continuity.

Enterprise-Grade Architecture

Unlike simple prompt templates, Acolytes for Claude Code is a complete development ecosystem with persistent memory, intelligent routing, and cross-agent coordination that scales with your project complexity.

Production-Ready Command System

With commands like /todo, /acolytes, /save, and /mcp, Acolytes for Claude Code provides a complete workflow management system that bridges sessions and maintains project continuity.

🎯 Acolytes for Claude Code vs Standard Claude Code

Feature Standard Claude Code Acolytes for Claude Code
Agents Single AI assistant 60 agents (55 specialized + 5 setup) + project Acolytes
Memory Session-only context Persistent SQLite + cross-session Jobs
Coordination Manual task switching Automatic Quest system with handoffs
Agent Selection User must specify Auto-routing based on context
Setup Manual configuration Auto-detection + 6-phase analysis
Project Learning Starts from zero Builds cumulative module knowledge
Error Recovery Manual intervention Automatic cross-agent coordination

🐛 Issues & Bug Reports

Found a bug or have an issue? Please report it at:

https://github.com/unmasSk/Acolytes/issues

Your feedback helps improve Acolytes for everyone!

📚 Documentation

Core System

Commands

Templates & Resources

🏆 Why Developers Choose Acolytes

🎯 Real Impact, Real Numbers

⚡ 26x faster API optimization (2.3s → 87ms)
🔍 50x faster code search with MCP code-index
🎯 Always up-to-date agents with Context7
🧠 14 memory types per project Acolyte (never forget anything)
🤝 Up to 10 specialists working in parallel
♾️ Infinite context through Jobs system
🚩 Zero coordination overhead with Quest System

🌟 Game-Changing Innovations

You used to: Write prompts, copy code, lose context, start over
Now you: Give one command, watch up to 10 specialists orchestrate the solution

You used to: Manually coordinate between frontend and backend
Now: Quest System automatically handles cross-domain coordination

You used to: Explain your project every new session
Now: Your project has a living, growing memory

💬 What Developers Say

"It's like having a senior dev team that never sleeps, never forgets, and never needs explanations twice."

"Quest System changed everything. My agents coordinate like never before."

"First time I've seen AI actually understand my entire codebase."

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.

Free for commercial and personal use.

⭐ The Future is Here

🎯 One request. Up to 10 specialists coordinated. Zero overhead.

Imagine typing "implement authentication" and watching as:

  • 🔐 Security architects design the system
  • 🗄️ Database experts create the schema
  • ⚙️ Backend engineers write the logic
  • 🎨 Frontend developers build the UI
  • 🚩 All coordinating through Quest System, automatically

❌ No more context switching.
❌ No more manual handoffs.
❌ No more starting from scratch.

Every decision, every piece of knowledge, every solution - preserved forever in your project's neural memory. 🧠

This isn't just an upgrade to Claude Code.
This is the emergence of autonomous development intelligence. 🤖


✨ Your code doesn't just get written.

🎭 It gets orchestrated.

pip install acolytes

🚀 Welcome to the revolution. ♾️

Metadata

Release files for acolytes 2.1.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for acolytes 2.1.2
File Size Uploaded Provenance
acolytes-2.1.2.tar.gz 1.4 MB Without provenance Details

Built distribution (wheel)

Table of built distributions (wheels) for acolytes 2.1.2
File Interpreter ABI Platform
acolytes-2.1.2-py3-none-any.whl Python 3 none any Details

Total release size: 2.9 MB

Release files / acolytes-2.1.2.tar.gz

Download URL acolytes-2.1.2.tar.gz
Size 1.4 MB
Tags Source
SHA-256 checksum
How to use checksums
0853d6a6ed1903a0f4aa6cf31a5e2f7bab7b951b59c5a2a583927db8f2493312
BLAKE2b-256 checksum
How to use checksums
d90868adc7acc490fb43814ae843d026577eb7cd46563bbf7fd9f88a70232490
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.1.0 CPython/3.13.3

Release files / acolytes-2.1.2-py3-none-any.whl

Download URL acolytes-2.1.2-py3-none-any.whl
Size 1.5 MB
Tags Python 3
SHA-256 checksum
How to use checksums
0dff091bb00bd361131c09e14a60f50582843670e3cf7cce3e918b4ea7aac46f
BLAKE2b-256 checksum
How to use checksums
5c833350246fee93b5b062223b9f03c746d302bea31bf11cdbc933cd683250cb
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.1.0 CPython/3.13.3

Security for acolytes 2.1.2

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page