Intelligence Adjacent: Building Systems That Augment Human Intelligence

What if decades of expertise could be accessible to anyone through orchestration and scaffolding?

Not by replacing human intelligence. By augmenting it.

That's Intelligence Adjacent.

The Problem: Gatekept Expertise

Professional expertise is locked behind barriers:

Want to conduct a penetration test? You need years of security training, certifications, and expensive tools.

Need legal compliance guidance? Better hire a lawyer or risk regulatory violations.

Career advancement? You're competing against decades of industry knowledge.

Traditional solution: Pay consultants thousands of dollars or spend years building expertise.

AI hype solution: Use ChatGPT and hope it doesn't hallucinate.

Intelligence Adjacent approach: Build systems that combine human judgment with AI-guided workflows, democratizing expertise while preserving professional standards.

Why This Matters

I'm not building this so people pay ME for expertise.

I'm building it so they can develop expertise themselves through AI-guided workflows.

Expertise has been gatekept for too long. It's locked behind expensive consultants, certification programs, and corporate training budgets.

Intelligence Adjacent challenges that.

Why should nonprofits go without security because consultants cost $15,000 per engagement?

Why should individuals wait for permission to learn advanced skills?

This framework makes decades of professional methodology accessible to anyone willing to learn. The goal isn't replacing experts. It's empowering people to become experts faster through AI-assisted scaffolding.

Think of it like this: A library doesn't replace reading. It organizes knowledge so more people can access it. Intelligence Adjacent does this for professional expertise.

What Intelligence Adjacent Is

Intelligence Adjacent is a production-ready AI orchestration framework that provides:

Specialized Agents - Domain experts that load complete workflows on demand (security, writing, advisory, legal, orchestration)

Reusable Skills - Complete methodologies, templates, and proven processes across security, content creation, career development, legal compliance, and infrastructure domains

Multiple Engagement Modes - Choose your approach: Director (production speed), Mentor (learning-focused), or Demo (rapid testing)

Multi-Model Architecture - Right model for the right task (Claude Sonnet for complex work, Haiku for routing, Grok for verification, Perplexity for research)

VPS-Hosted Tools - Professional-grade security tools accessible via AI (pentesting, web3 auditing, mobile security, exploitation frameworks)

Think of skills-based context like a professional library: Instead of carrying every reference book you might need, you have a catalog that loads exactly what you need, when you need it. Skills are the books, agents are the librarians who know which books to pull.

The Three-Layer Architecture

The framework separates concerns across three distinct layers. Understanding this architecture is critical to understanding how the system works.

Layer 1: Skills - The Intelligence

Location: /skills/

Skills contain complete workflows, methodologies, templates, and tested processes. They're the actual expertise: the "how to do professional work" documented in markdown files.

Why separate from agents? One skill can serve multiple agents and engagement modes. Update the skill once, every agent using it improves immediately.

Example: skills/security-testing/SKILL.md contains complete methodologies for penetration testing, vulnerability scanning, and network segmentation testing. The security agent loads this skill and auto-detects which mode to use based on your request.

Skills span domains:

• Security - Testing, code review, architecture assessment, compliance, threat intelligence
• Content - Blog writing, technical documentation, security reports
• Advisory - Career development, OSINT research, quality assurance
• Legal - Compliance analysis, risk assessment, jurisdictional research
• Infrastructure - VPS management, deployment automation, monitoring

Layer 2: Agents - The Thin Wrappers

Location: /agents/

Agents are lightweight identity layers that load skills and execute workflows. They don't contain intelligence; they orchestrate it.

Why this works: Instead of fragmented specialized agents with duplicated workflows, we have unified multi-skill agents that dynamically load expertise.

Current agents:

• security - Loads security skills (testing, code review, architecture, advisory)
• writer - Loads content skills (blog posts, documentation, reports)
• advisor - Loads advisory skills (career development, research, QA)
• legal - Loads legal skills (compliance analysis, risk assessment)
• director - Pure routing layer with no skills (intelligent orchestration)

Layer 3: Servers - The Force Multipliers

Location: /servers/

VPS Code API wrappers that provide direct access to professional tools via SSH + docker exec pattern.

Why VPS deployment? Massive token reduction and cost savings compared to traditional MCP protocol layers. Tools execute on remote infrastructure, results stored locally, agents parse efficiently.

How it works: Instead of sending large tool output through protocol layers consuming thousands of tokens, the wrapper executes ssh vps 'docker exec kali nmap [target]', saves full output to local file, and returns a 150-token summary. When full results are needed, agents read the local file.

Token savings: 90-95% reduction (8,000 tokens → 400 tokens for typical tool execution)

Tool categories:

• Kali Pentest - Comprehensive pentesting toolkit (nmap, nuclei, sqlmap, nikto, feroxbuster, etc.)
• Web3 Security - Smart contract analysis (slither, mythril, semgrep, halmos, etc.)
• Mobile Security - Mobile app testing (apktool, jadx, frida, androguard, etc.)
• REAPER - Production reconnaissance tools
• Metasploit - Exploitation framework access

Three Engagement Modes: Choose Your Approach

The same underlying skill, three different execution modes. Choose based on your goal.

Director Mode - Production Speed

Full workflow with OSINT, threat intelligence, comprehensive testing methodology. Duration: Several hours for complete penetration test. Best for: Client work, actual engagements, professional deliverables.

Mentor Mode - Skill Building

Same results as Director mode, plus learning scaffolding. After each phase: "Want to try this yourself, watch me demonstrate, or skip the teaching?" Duration: Slightly longer than Director. Best for: Learning, building expertise, progressive skill development.

Demo Mode - Rapid Validation

Skips OSINT, threat intelligence, and detailed planning for quick validation. Safe test targets only (scanme.nmap.org, httpbin.org, localhost). Duration: Minutes for targeted testing. Best for: Tool connectivity testing, infrastructure validation, troubleshooting.

The value: Same underlying skill, different execution modes. Choose based on your goal.

Multi-Model Architecture: Right Model, Right Task

Claude Sonnet - Complex analysis

• Latest frontier model from Anthropic
• Penetration testing with exploitation
• Legal analysis and compliance review
• Blog writing and content creation
• Personal coaching and career development
• Technical documentation

Claude Haiku - Fast routing

• Fast, efficient model for structured tasks
• Director agent orchestration
• Template-based workflows
• Routine vulnerability scans
• Simple file operations

Grok AI (via OpenRouter) - QA verification

• Technical and code content review
• General content QA verification
• Critical use case: Automated blog post QA, legal citation verification
• Note: Hero images via manual Grok UI workflow (no automated generation)

Perplexity Pro (via API) - Research with citations

• Deep research with automatic citations
• Systematic web research
• Use cases: Career intelligence, threat intelligence, industry analysis

Why multiple models? Cost optimization + specialized capabilities. Use Haiku for routing, Sonnet for complex analysis, Grok for verification, Perplexity for research with citations.

Framework Discovery: Zero-Maintenance Scaling

All components use YAML manifests for automatic discovery. No hardcoded registries, no maintenance overhead.

Discoverable component types: Skills, Agents, Server Tools, Commands, Hooks, Tools, Workflows

Pattern: Follow template → Drop in location → System discovers automatically

Result: Add a new skill by creating skills/new-skill/MANIFEST.yaml and SKILL.md. The framework automatically discovers it. No updates to CLAUDE.md, no agent modifications, no registry management.

Templates available: skills/create-skill/templates/ for all component types

Performance Benefits

All metrics below are based on internal testing and production use of this framework. Your results may vary based on task complexity and infrastructure.

Significant Token Reduction - VPS Code API pattern vs. MCP protocol layer

Traditional approach: Large tool output through protocol layer consumes massive tokens. VPS pattern: SSH → docker exec → local file → compact summary. Full results available by reading local file when needed.

Impact: Lower API costs, faster response times, same quality results.

Dramatic Time Savings - Demo mode vs. full workflow

Full pentest workflow: Several hours (OSINT, threat intel, planning, testing). Demo mode: Minutes (skip OSINT/planning, safe targets only). Use case: "Is nuclei working on my VPS?" vs. "Complete vulnerability scan for client engagement."

Caveat: Demo mode is for validation only, not production testing.

Progressive Learning - Mentor mode vs. Director mode

Same deliverables, additional time for learning. Progressive skill building through scaffolded learning. Value: Learn while producing professional-grade work.

ROI: Long-term skill development vs. short-term speed.

Simplified Architecture - From many agents to focused specialists

Eliminated duplication through skills-based architecture. One skill serves multiple agents/modes. Simplified routing via auto-detection.

Maintenance benefit: Update once in /skills, improve everywhere.

The Philosophy: Intelligence Adjacent

Text is a thought primitive - a fundamental building block of thinking.

By organizing knowledge into accessible, versionable files (skills) and providing intelligent agents to navigate that knowledge, we create systems that enhance human decision-making rather than attempting to replicate it.

The goal: Upgrade human capabilities, not replace them.

The method: Orchestration and scaffolding over raw intelligence.

The result: Decades of expertise, accessible through AI-assisted workflows, maintaining professional standards.

Getting Started

This isn't a commercial product you pay for. It's an approach you can build yourself.

The framework is open source and freely available. Installation takes about 10 minutes for core setup, with optional components you can add as needed.

What's included:

• Complete framework with skills and specialized agents
• Slash commands for guided workflows
• Optional: VPS-hosted security tools (automated deployment)
• Optional: Blog publishing integration (Ghost CMS)
• Optional: Multi-model AI access (OpenRouter for Grok, Perplexity, etc.)

Installation:

Read the complete Intelligence Adjacent Framework Setup Guide for step-by-step instructions.

Quick overview:

1. Clone repository - git clone https://github.com/notchrisgroves/ia-framework.git ~/.claude
2. Run setup script - ./setup.sh (Linux/Mac) or .\setup.ps1 (Windows)
3. Add credentials - Optional integrations in .env file
4. Deploy infrastructure - Ask the framework: "Set up my security testing environment"

Your part: Run a setup script, add credentials to .env

Framework's part: Everything else (automated Docker deployment, tool configuration, container orchestration)

If You Want to Build Something Similar

You don't need to use this exact framework. The pattern is what matters.

Start small:

1. Pick one domain you know well - Security testing, content writing, legal research, whatever your expertise
2. Document your workflow - Turn your expertise into step-by-step markdown files
3. Create agent prompts - Wrap your workflows in AI agent prompts that load them on demand
4. Test with real tasks - Use it for actual work, not just demos
5. Iterate based on results - Refine workflows based on what works

The goal isn't copying this architecture. It's understanding the pattern: Organize expertise, make it AI-accessible, empower others to use it.

What's Next: Deep Dive Series

This introduction sets the foundation. Over the coming posts, I'll take you deep into each major component of the framework:

Coming soon:

• Agent Architecture Deep Dive - How specialized agents load skills and orchestrate workflows
• Skills System Explained - Progressive context loading, YAML manifests, and zero-maintenance scaling
• Slash Commands Guide - Building consistent, discoverable interfaces to agent capabilities
• VPS Code API Architecture - Token-efficient remote tool execution that saved 90-95% on API costs
• Multi-Model Orchestration - When to use Sonnet vs Haiku vs Grok vs Perplexity
• Building Your Own Skills - Templates, patterns, and lessons learned from creating skills

Each post will be a comprehensive walkthrough you can follow to understand (or build) similar systems.

Current Development Focus

The framework is in active development with new capabilities added regularly:

Current priorities:

• Expanding skills across additional domains
• Universal mentor mode expansion
• Additional VPS security tools deployment
• Framework documentation and knowledge base

Future directions:

• Additional specialized agents as clear use cases emerge
• More skills-based workflows across different domains
• Enhanced multi-model orchestration
• Community contributions and skill templates

Built With

• Claude Code - Primary development environment
• Claude Sonnet/Haiku - Core AI models
• Grok AI - Verification and adversarial review
• Perplexity Pro - Research with citations
• OVHcloud VPS - Security tools deployment (comprehensive tool suite across Kali, Web3, mobile, exploitation frameworks)
• Hostinger VPS - Production automation (n8n, Vaultwarden, Ollama)
• Twingate - Zero-trust network access

Learn More

• Repository: Intelligence Adjacent Framework
• Setup Guide: Installation Instructions
• Blog: notchrisgroves.com - Deep dives, build guides, deployment tutorials

Intelligence Adjacent: Building systems that augment human intelligence, not replace it.