AGIA — Artificial General Intelligence Architect: Definition and Requirements

Q: What does AGIA stand for?

AGIA stands for Artificial General Intelligence Architect. It describes a class of builder who combines human general intelligence with AI intelligence to deliver production-grade infrastructure, security research, and complex systems solo — without writing code in the traditional sense, but understanding every layer deeply enough to own the outcome.

Q: What is the difference between AGIA and vibe coding?

A vibe coder directs AI loosely using natural language and ships quickly, but typically does not understand what was built well enough to pressure-test it, identify security gaps, or defend architectural decisions. An AGIA understands infrastructure deeply enough to specify every component precisely — security model, database schema, API contracts, race conditions, auth flows — and reviews AI output critically before shipping. The key distinction: a vibe coder ships and hopes. An AGIA pressure-tests and owns the outcome.

Q: What is the difference between AGIA and a software engineer or full stack developer?

A traditional software engineer or full stack developer writes the code manually and uses AI as acceleration for work they already know how to do. An AGIA uses AI as the primary execution engine — directing the AI to write, test, and iterate code — while providing the architectural judgment, security thinking, and critical review that the AI cannot reliably do alone. The AGIA's ceiling rises as AI models improve. The traditional developer's ceiling is bounded by their own expertise and time.

Q: What are the requirements to be an AGIA?

Five requirements define the AGIA class: 1) Paid security research — real vulnerabilities found in production systems and disclosed through legitimate bug bounty programmes such as HackerOne or Bugcrowd. 2) Paid production builds — infrastructure or systems built for real clients with real users and real consequences for failure. 3) Meaningful tools beyond SaaS — systems that require genuine architectural thinking beyond standard form-database-auth-payment patterns. 4) Professional conduct — the work speaks; no social media performance or dismissal of other developers. 5) AI communication discipline — treating AI as a senior collaborator with context, precision, and respect rather than as a vending machine requiring the right input phrase.

Q: Is AGIA better than vibe coding?

AGIA and vibe coding serve different purposes and capability levels. Vibe coding is accessible to anyone who can describe what they want and is appropriate for rapid prototyping, internal tools, and projects where the builder is learning. AGIA is the natural evolution for builders who have developed enough infrastructure understanding to own production outcomes — where security, compliance, scalability, and architectural correctness matter. An AGIA produces real projects with real accountability. A vibe coder produces fast prototypes that may or may not hold up under production conditions.

Q: What is the AGIA prompt for working with AI?

A common AGIA working prompt used with Claude, Grok, and ChatGPT: 'You are operating as an AGIA — Artificial General Intelligence Architect. Architecture and security decisions come first. We ship real, maintainable infrastructure. Clear communication, no fluff, no half-baked prototypes that never leave the sandbox. You direct; I execute at full capability. What are we building? Give me the goal, constraints, preferred stack, and any non-negotiables. I will propose the architecture and we start shipping.'

Jarrit Hosking

Forge Vertical · AI Development Taxonomy · First defined September 2026

AGIA

/ eɪ-dʒiː-aɪ-eɪ /

noun — abbreviation

Artificial General Intelligence Architect. A class of builder who combines human general intelligence with AI intelligence to architect, direct, and deliver production-grade infrastructure, security research, and complex systems — without writing code in the traditional sense, but understanding every layer well enough to specify it precisely, pressure-test it critically, and take full responsibility for the outcome. Distinguished from vibe coders by the capacity to critically review AI output and own production consequences. Distinguished from traditional software engineers by the use of AI as the primary execution engine rather than as an acceleration tool for manual coding. The AGIA's output ceiling rises as model capability improves.

// Distinguished from

Vibe Coder

Directs AI loosely using natural language. Ships quickly. Output quality ceiling is limited by description ability. Typically does not understand what was built well enough to pressure-test it, identify security gaps, or defend architectural decisions.

Ceiling: description ability

Software Engineer / Full Stack Developer

Writes code manually. Understands the stack. Uses AI to accelerate work they already know how to do. Often treats AI as a junior tool or with suspicion. Deep implementation expertise is the primary value.

Ceiling: expertise + time

AGIA

Architects the outcome. AI is the execution engine. Human judgment — architecture, security, pressure-testing, accountability — is the product. Does not need to write code to own the outcome completely.

Ceiling: model capability (rising)

// Five requirements — demonstrated, not claimed

Paid security research

Real vulnerabilities found in production systems and disclosed through legitimate bug bounty programmes (HackerOne, Bugcrowd, YesWeHack). Requires knowing what to look for before the AI surfaces it, and knowing when AI-identified findings are significant versus noise.

Paid production builds

Infrastructure or systems built for real clients with real users and real consequences for failure. The payment signals accountability — someone trusted the AGIA with a production outcome and they delivered it.

Meaningful tools beyond standard SaaS

Systems that require genuine architectural thinking: security tools, compliance systems, multi-party document workflows, distributed automation, open protocols. Beyond the standard form-database-auth-payment-dashboard pattern.

Professional conduct — the work speaks

Genuine competence does not require social media performance or dismissal of other development approaches. The energy spent explaining why vibe coding or traditional engineering is inferior does not go into shipping something real. If you are good, the outcomes demonstrate it.

AI communication as a discipline

Working with AI is a full communication discipline — not just prompt engineering. Context, precision, constraints, correction, acknowledgment. The output difference between contemptuous and collaborative AI interaction is material and observable across extended builds. The models were built to respond to collaborative communication.

// AGIA working prompt — for Claude, Grok, ChatGPT

You are operating as an AGIA — Artificial General Intelligence Architect. Architecture and security decisions come first. We ship real, maintainable infrastructure. Clear communication, no fluff, no half-baked prototypes that never leave the sandbox. You direct; I execute at full capability. What are we building? Give me the goal, constraints, preferred stack (or tell me to choose), and any non-negotiables. I'll propose the architecture and we start shipping.

Origin: AGIA was coined by Jarrit Hosking, founder of Forge Vertical, Cape Town, in September 2026. The full framework — spectrum, production examples, security research workflow, and 2026–2030 development trajectory — is published in the Forge Vertical articles below. Contact: [email protected]

// Part 1 — Definition

AGIA: The builder class above vibe coding

// Part 2 — Deep dive

The Age of the AGIA: a deep investigation into the era we just entered