HeliOS Studio

Personal AI R&D Lab - One-person AI-powered startup studio for discovery, research, architecture, prototyping, and deployment

Overview

HeliOS Studio is a complete AI-powered development ecosystem that enables a solo technologist to operate like a small R&D lab or startup studio. By orchestrating multiple AI tools, local models, and automation platforms, you can:

• Discover business opportunities through systematic research
• Build and maintain a persistent knowledge base
• Design system architectures with AI assistance
• Generate working software prototypes rapidly
• Deploy and operate projects with minimal manual overhead

Architecture

┌─────────────┐
│  Discovery  │  Perplexity AI → NotebookLM
└──────┬──────┘
       │
┌──────▼──────┐
│  Knowledge  │  NotebookLM + GitHub Knowledge Repos + Obsidian
└──────┬──────┘
       │
┌──────▼──────┐
│  Reasoning  │  Claude + Claude Code + Local LLMs
└──────┬──────┘
       │
┌──────▼──────┐
│   Coding    │  Cursor / VS Code + Copilot
└──────┬──────┘
       │
┌──────▼──────┐
│ Automation  │  n8n + MCP Tools + GitHub Actions
└──────┬──────┘
       │
┌──────▼──────┐
│    Infra    │  Docker + Proxmox + GitHub + Cloud
└─────────────┘

New: Tools now integrate directly via MCP servers and APIs—Claude can build n8n workflows, query NotebookLM notebooks, and manage GitHub Projects conversationally. See Integrations Guide.

Quick Start

1. Architecture Overview - Understand the system design
2. Tool Stack - Recommended tools for each layer
3. Integrations - Direct tool connections via MCP and APIs ✨ NEW
4. Setup Guide - Step-by-step installation
5. Workflows - End-to-end examples
6. Security - Identity separation and data protection

Core Principles

• AI-first: Use AI for all code generation, architecture, and research
• GitHub-centric: Single source of truth for all work
• Local + Cloud: Balance cost, privacy, and capability
• Automated: Minimize manual handoffs between tools via MCP and webhooks
• Secure: Strong compartmentalization and data policies

Documentation

Getting Started

• Architecture Overview
• Tool Stack Recommendations
• Tool Integrations & MCP Servers ✨ NEW
• Complete Setup Guide
• Hardware Recommendations

Operations

• Example Workflows
• Project Management with GitHub Projects ✨ NEW
• Automation Patterns
• Security & Identity Management
• Secure-by-Design Recommendations ✨ NEW

Reference

• MCP Integration Guide
• Local LLM Strategy
• IDE Comparison
• Future Roadmap

Key Integrations

MCP Servers (March 2026)

Claude Desktop can now directly:

• Build n8n workflows conversationally (n8n MCP server)
• Query NotebookLM notebooks for research (unofficial MCP server)
• Read/write Obsidian vault for personal knowledge (Obsidian MCP server)
• Manage GitHub repos, issues, PRs (official GitHub MCP server)

Automation

n8n can now:

• Call Perplexity API for scheduled research queries
• Export NotebookLM notebooks via Apify actor
• Receive Cursor webhooks when agents complete
• Update GitHub Projects via GraphQL API

Result: Entire workflows from research → architecture → prototype → experiment can run with minimal manual intervention.

See Integrations Guide for setup instructions.

Project Management

HeliOS-Studio uses GitHub Projects v2 as the central control tower:

• Native GitHub integration (zero sync)
• Claude can manage via GitHub MCP
• n8n automation for status updates
• Custom fields: AI Agent, Phase, Health, Cost Estimate
• Multiple views: Opportunities Board, Active Sprints, Timeline, Agent Dashboard

See Project Management Guide for setup and automation workflows.

Project Ideas

This stack enables rapid development of:

• Cybersecurity tools: MCP security scanners, prompt-injection gateways, threat intelligence aggregators
• Developer tooling: Studio-in-a-box templates, local model evaluators, code quality agents
• Automation systems: Homelab orchestrators, experiment runners, opportunity miners
• AI agents: Multi-model routers, research assistants, workflow coordinators

Ecosystem Monitoring

⚠️ Important: The AI tooling landscape evolves rapidly. As of March 2026, there are 8,590+ MCP servers and new integrations launch weekly.

Monthly review checklist:

• Check PulseMCP.com/servers for new MCP servers
• Review n8n integrations page for new native nodes
• Check Cursor, Claude, NotebookLM release notes
• Test existing MCP servers still work
• Update documentation with new capabilities

Automated: An n8n workflow creates a GitHub issue with this checklist on the first Monday of each month.

See Integrations Guide - Maintenance Strategy for details.

Status

• Architecture design
• Tool stack research
• Security framework
• Direct integrations research ✨ NEW
• Project management strategy ✨ NEW
• Phase 1 setup (core tools)
• Phase 2 setup (MCP + automation)
• Phase 3 setup (knowledge pipeline)
• GitHub Project creation
• First prototype project

Contributing

This is a personal R&D lab, but if you're building something similar:

1. Fork this repo as a template for your own studio
2. Adapt the tool stack to your preferences
3. Share your learnings via issues or PRs
4. Join the conversation about AI-augmented development

License

MIT

Author

Built by a solo technologist with 25+ years in IT and cybersecurity, exploring the frontier of AI-augmented software development.

Meta: This studio documentation was researched, designed, and written with assistance from Perplexity AI, Claude, and the ecosystem of tools it describes. 🤖