Skip to content

DafnckStudio/DafnckMachine-v3.1

BranchesTags

Repository files navigation

$$$$$$$\ $$$$$$\ $$$$$$$$\ $$\ $$\ $$$$$$\ $$\ $$\ $$ __$$\ $$ __$$\ $$ _|$$$\ $$ |$$ $$\ $$ | $$ | $$ | $$ |$$ / $$ |$$ | $$$$\ $$ |$$ / _|$$ |$$ / $$ | $$ |$$$$$$$$ |$$$$$\ $$ $$$$ |$$ | $$$$$ /
$$ | $$ |$$ $$ |$$ | $$ $$$$ |$$ | $$ $$<
$$ | $$ |$$ | $$ |$$ | $$ |$$$ |$$ | $$\ $$ |$$\
$$$$$$$ |$$ | $$ |$$ | $$ | $$ |$$$$$$ |$$ | $$\ _/ _| _|_| _| __| _/ _| _|

🚀 DafnckMachine - Agentic Coding Framework

Automate Your Vision into Realit - Build anything ! Transforming software / app / saas / game development with spec-driven, AI-powered agentic workflows.

Build Status License Version Contributions Welcome

DafnckMachine-V3.1

Autonomous AI Workflow Orchestration Platform

🚀 What is DafnckMachine?

DafnckMachine is a next-generation, agent-driven workflow orchestration platform designed for AI-augmented software development. It enables both AI agents and human users to collaborate on complex projects, automating everything from requirements gathering to deployment and continuous improvement.

  • Multi-agent system: 67+ specialized agents for every phase of the SDLC.
  • Smart Brain System: DNA, STEP, GENESIS JSONs for state, workflow, and learning.
  • Task Master Integration: Full-featured task breakdown, tracking, and automation.
  • Performance-optimized: Lightweight configs, caching, and real-time state sync.
  • Cursor/RooCode Ready: Native support for agent invocation and workflow navigation.

How to Start from a Small Idea or Blank Slate

You can begin your DafnckMachine project with just a rough idea, a few features, or even nothing at all!

  1. Open and Edit @Project.md:

    • Jot down any initial thoughts, goals, or features you have in mind—even if it's just a sentence or a few bullet points.
    • This file will serve as the seed for your project vision and requirements.
    • Don't worry about completeness; you can always update it later.

  2. Save @Project.md

    • The more detail you provide, the less the agents will need to ask. But you can start with as little as a single idea.

  3. Prompt the System (Recommended):

    • In Cursor or RooCode chat, type @uber-orchestrator-agent and describe your idea, or simply ask "Help me start a new project."
    • The orchestrator agent will analyze your @Project.md, ask for any missing info, and guide you through the next steps.

  4. Start the Workflow (Optional):

    • For a full environment setup, begin with [00_Project_Initialization.md](01_Machine/01_Workflow/Phase 0 : Project Setup/00_Project_Initialization.md).
    • For user context and requirements gathering, start with [P01-S01-T01-User-Profile-Development.md](01_Machine/01_Workflow/Phase 1: Initial User Input & Project Inception/01_User_Briefing/P01-S01-T01-User-Profile-Development.md).

  5. Let the Agents Guide You:

    • The agents will read your @Project.md and only ask for missing or unclear information.
    • They will guide you through the rest of the process, step by step, adapting to what you have already provided.

Tip: If you're unsure where to start, just type @uber-orchestrator-agent in chat and say "Start my project"—the system will handle the rest!

The workflow is adaptive—add as much or as little as you want to @Project.md. The system will fill in the gaps and help you clarify your vision as you go!

🏗️ System Architecture

DafnckMachine-V3.1/
├── 01_Machine/           # The Engine (How to execute)
│   ├── 01_Workflow/      # Step-by-step execution plans
│   ├── 02_Agents/        # Agent definitions and capabilities  
│   ├── 03_Brain/         # Intelligence system (DNA, STEP, GENESIS)
│   └── 04_Documentation/ # System documentation
├── 02_Vision/            # The Strategy (What to build)
│   ├── Project goals and vision
│   ├── Strategic direction
│   └── High-level requirements
└── 03_Project/           # The Output (What gets built)
    ├── Actual project files
    ├── Generated code
    └── Implementation artifacts
  • DNA.json: Agent registry, capabilities, and communication protocols
  • STEP.json: Execution engine with task mapping and state management
  • GENESIS.json: Adaptive configuration and learning system
  • AGENT_INTERFACE.json: ⚡ Performance-optimized lightweight configs

⚡ Quick Start

For AI Agents (Cursor/RooCode)

  1. Open the current workflow file:
    01_Machine/01_Workflow/{Phase}/[Step].md
  2. Read the Agent Context at the top for instructions and config.
  3. Use @agent-name in Cursor/RooCode chat to invoke a specific agent (see below).
  4. Follow the numbered tasks (1.1, 1.2, etc.) in the workflow file.
  5. Output results to 03_Project/{step_outputs}/.
  6. Update progress in the workflow file and state JSONs.
  7. Navigate to the next step using the provided links.

For Human Users (CLI/Manual)

  1. Review the architecture:
    01_Machine/04_Documentation/01_System/Project_Structure_Integration.md
  2. Check the Agent Operations Manual:
    01_Machine/04_Documentation/01_System/Agent_Operations_Manual.md
  3. Monitor progress in 01_Machine/01_Workflow/ files.
  4. Review outputs in 03_Project/ directories.
  5. Use Task Master CLI for task management (see below).

🤖 Using with Cursor/RooCode

  • Invoke agents in chat with @agent-name (e.g., @coding-agent, @uber-orchestrator-agent).
  • Agents collaborate: Mention multiple agents for complex tasks.
  • Agent context: Each workflow file specifies the responsible agent and their capabilities.
  • Workflow navigation:
    • Use the links in workflow files to move between steps.
    • Progress and state are tracked in 01_Machine/03_Brain/Step.json and workflow_state.json.
  • Best practices:
    • Be specific in your requests to agents.
    • Reference the correct agent for each phase (see agent list below).
    • Combine agents for multi-domain tasks.
    • Provide context for better results.

Example agent invocations:

@uber-orchestrator-agent Planifie la roadmap de ce projet
@coding-agent Implémente une API REST pour la gestion des tâches
@ui-designer-agent Propose un design moderne pour le dashboard

🛠️ Task Master Workflow (MCP & CLI)

DafnckMachine uses Task Master for all task, subtask, and workflow management.

MCP Server (Recommended for Cursor/RooCode)

  • Use integrated tools (MCP) for best performance and error handling.
  • Key commands:
    • get_tasks / task-master list
    • next_task / task-master next
    • expand_task / task-master expand --id=<id>
    • set_task_status / task-master set-status --id=<id> --status=done
    • add_task, add_subtask, update_task, update_subtask, etc.

CLI Usage

  • Install: npm install -g task-master-ai
  • Or use locally: npx task-master-ai ...
  • See .roo/rules/taskmaster.md for full command reference.

Workflow Example

  1. task-master init — Initialize project structure.
  2. task-master parse-prd --input='scripts/prd.txt' — Generate initial tasks.
  3. task-master list — View all tasks.
  4. task-master next — See the next actionable task.
  5. task-master expand --id=1 --force --research — Break down complex tasks.
  6. task-master set-status --id=1.1 --status=done — Mark subtasks as complete.

🧠 Agent System & Collaboration

  • 67+ specialized agents for every phase of the SDLC.
  • Invoke with @agent-name in Cursor/RooCode or reference in workflow files.
  • Collaboration: Agents can be combined for multi-domain tasks.
  • Agent registry: See 01_Machine/02_Agents/ and DNA.json for full list.

Key agent categories:

  • Orchestration: @uber-orchestrator-agent, @development-orchestrator-agent
  • Planning: @task-planning-agent, @prd-architect-agent
  • Development: @coding-agent, @system-architect-agent, @code-reviewer-agent
  • Design/UX: @ui-designer-agent, @ux-researcher-agent, @design-system-agent
  • Testing: @test-orchestrator-agent, @functional-tester-agent, @security-auditor-agent
  • Documentation: @scribe-agent, @documentation-agent, @elicitation-agent
  • DevOps: @devops-agent, @adaptive-deployment-strategist-agent
  • Analytics/Marketing: @analytics-setup-agent, @marketing-strategy-orchestrator

See:

  • 01_Machine/04_Documentation/01_System/AGENTS_README.md
  • 01_Machine/04_Documentation/01_System/AGENT_GUIDE.md
  • .roo/rules/dev_workflow.md for agent workflow best practices.

📁 Directory & Documentation

  • 01_Machine/01_Workflow/: Step-by-step execution plans (primary workspace)
  • 01_Machine/02_Agents/: Agent definitions and capabilities
  • 01_Machine/03_Brain/: Core system state (DNA, STEP, GENESIS, AGENT_INTERFACE)
  • 01_Machine/04_Documentation/: System docs, templates, and guides
  • 02_Vision/: Project vision, strategy, and high-level requirements
  • 03_Project/: All generated outputs and deliverables

Key docs:

  • 01_Machine/04_Documentation/01_System/Agent_Operations_Manual.md
  • 01_Machine/04_Documentation/01_System/Project_Structure_Integration.md
  • 01_Machine/04_Documentation/01_System/Template-Step-Structure.md
  • .roo/rules/dev_workflow.md
  • .roo/rules/taskmaster.md
  • .roo/rules/self_improve.md

🎯 Best Practices

  • Start with workflow files — All instructions are embedded.
  • Use lightweight configs — AGENT_INTERFACE.json for most operations.
  • Follow numbered task structure — 1.1, 1.2, 2.1, etc.
  • Update progress in real-time — Keep checklists current.
  • Reference vision strategically — Only when needed for decisions.
  • Output to structured directories — Follow 03_Project/ organization.
  • Iterate and improve rules — See .roo/rules/self_improve.md.

🔧 Troubleshooting & Continuous Improvement

  • Agent not responding? — Check Agent Context in workflow file and agent registry.
  • Performance issues? — Use lightweight configs, check for caching.
  • Task Master errors? — Ensure API keys are set in .env or .cursor/mcp.json.
  • Rule evolution — Add/update rules in .roo/rules/ as new patterns emerge.
  • Validate agents — Use unified agent validator scripts in 01_Machine/03_Brain/Agents-Check/.

📈 System Evolution

  • GENESIS system learns from agent performance and usage.
  • Automatic optimization and real-time adaptation.
  • Performance feedback loops for system enhancement.

📝 Contributing

  • Add new agents in 01_Machine/02_Agents/ and update DNA.json.
  • Document new steps using Template-Step-Structure.md.
  • Improve rules in .roo/rules/ as new best practices emerge.
  • See: 01_Machine/04_Documentation/01_System/00_Documentation_Index.md for step documentation.

🧪 Agent Validation & System Health

What is unified_agent_validator.py and unified_agent_validator.sh?

  • Location: 01_Machine/03_Brain/Agents-Check/
  • Purpose: These scripts validate all agent definitions, check for errors, repair common issues, and ensure the system is ready for use.
  • Features:
    • Validate agent JSON structure and required fields
    • Auto-repair common format and reference issues
    • Test agent loading and system initialization
    • Generate comprehensive health reports
    • Sync agent definitions to .roomodes (RooCode) and .cursorrules (Cursor)
  • How to use:
    • Run python3 unified_agent_validator.py (for full-featured validation, repair, and reporting)
    • Or run bash unified_agent_validator.sh (for shell-based validation and repair)
    • Use menu options or CLI flags for repair, sync, and report generation
    • See script comments and help output for advanced usage

⚙️ MCP Server Configuration

DafnckMachine uses MCP servers for advanced agent and workflow integration (Cursor, RooCode, Task Master, etc).

How to Set Up MCP Servers

  1. Locate the template config:
    • .roo/mcp-template.json (for RooCode)
    • .cursor/mcp-template.json (for Cursor)
  2. Copy and rename:
    • Remove -template from the filename (e.g., .roo/mcp.json)
  3. Edit the config:
    • Fill in your API keys for each service (e.g., Anthropic, Perplexity, Supabase, GitHub, Stripe, etc.)
    • Update any paths or user IDs as needed
    • Do not share your API secrets publicly
  4. Supported servers (examples):
    • taskmaster-ai (Task Master integration)
    • supabase (database)
    • github (repo management)
    • firebase, playwright, puppeteer, stripe, context7, perplexity-mcp, shadcn, langgraph, three-devtools, elizaOS, memory, everything, railway, Framelink Figma MCP, A2A Docs, Pheromind Docs, n8n Docs, chakra-ui Docs, @21st-dev/magic, etc.
    • Add or remove servers as needed for your project
  5. Activate the config:
    • The system will use .roo/mcp.json or .cursor/mcp.json automatically if present
    • Restart your MCP server or reload your workspace if needed

Note:

  • Never commit your API secrets to public repositories.
  • See the template files for all available server options and required fields.
  • For more details, see the comments in the template files and the main README.

Last Updated: 2025-06-05
Support: See 01_Machine/04_Documentation/
Brain Status: Optimized for performance and usability
Performance: Lightweight interface active

About

Autonomous Agentic Coding Framework

Resources

Readme

License

MIT license
Activity

Stars

58 stars

Watchers

2 watching

Forks

35 forks
Report repository

Releases

No releases published

Packages

No packages published

Languages