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.
🌍 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
/setupPhase 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:
- 📋 INSTALLATION.md - Complete setup guide with proper command sequences
- 📖 HOW-TO.md - How to use the Quest system and communicate with agents
- 💾 SAVE.md - Memory system and session persistence
- 🚀 QUEST.md - Understanding the revolutionary Quest coordination
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.pythonor@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)
agents_catalog- Directory of all 60 agents with routing rulesagents_memory- Deep knowledge storage for each agent/modulesessions- Conversation history with context preservationmessages- Individual message tracking with metadatatool_logs- Execution history for optimizationtodos- Task management synced with Claudeacolyte_quests- Quest coordination dataacolyte_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, TODOsstructure- Code organization: files, classes, functions, APIspatterns- Best practices: conventions, design patterns, anti-patternsinterfaces- Module exports: public APIs, events, contractsdependencies- Connections: internal/external deps, servicesschemas- Data models: entities, validation, transformationsquality- Code health: tests, coverage, security analysisoperations- DevOps: deployment, monitoring, CI/CD, migrationscontext- Business logic: decisions, history, roadmapdomain- Specialized: ML models, GraphQL, i18nsecurity- Security profile: permissions, compliance, vulnerabilitieserrors- Error handling: common failures, recovery proceduresperformance- Optimization: bottlenecks, caching, scalinghistory- 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:
@coordinator.security→ Recommends OAuth2 + JWT hybrid approach@service.auth→ Implements OAuth flow with passport.js@database.postgres→ Creates users, sessions, oauth_providers tables@backend.nodejs→ Integrates middleware and endpoints@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:
@coordinator.migration→ Creates migration strategy@database.mongodb+@database.postgres→ Schema mapping- Data type mapping coordinated automatically
@backend.nodejs→ Updates all queries and models@ops.monitoring→ Sets up migration monitoring- 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
- Agent Catalog - All 60 agents with capabilities
- Agent Routing Rules - 17-rule disambiguation system
Commands
- Setup Command - 6-phase project initialization
- Commit Command - Smart commits with auto-versioning
- PR Command - Pull request automation
- Acolytes Command - Process coordination tasks (IN DEVELOPMENT)
- Save Command - Session persistence
Templates & Resources
- CLAUDE.md Template - Main project configuration
- Acolytes Template - Acolyte agent creation
- Architecture Template - System design docs
- Infrastructure Template - DevOps setup
- Roadmap Template - Project planning
- Tech Stack Template - Technology choices
- Vision Template - Project goals & mission
🏆 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. 🤖
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)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| acolytes-2.1.2.tar.gz | 1.4 MB | Without provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| 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.