petertubin-droid

Archie — Development skill for Claude Code

Development community

ARCHIE Native Intelligence Engine — an independent, provider-agnostic cognitive engine (15 systems) with its own edge-function runtime, knowledge subsystems and PWA.

How to install Archie

This entry records only its repository, not the path inside it, so there is no exact command to give. Open petertubin-droid/ARCHIE and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Archie does

ARCHIE Native Intelligence Engine — an independent, provider-agnostic cognitive engine (15 systems) with its own edge-function runtime, knowledge subsystems and PWA. Extracted from FRELUX; zero dependence on Gemini/OpenAI/Claude.

Alternatives in Development

Dive — Dive is an open-source MCP Host Desktop Application that seamlessly integrates with any LLMs supporting functi 1.8k ★
Codenotch — A macOS app that pins usage limits from Claude Code, Cursor, Codex, and Antigravity to a screen edge 1.3k ★
Claude Code Sub Agent Collective — 🧠 Context Engineering Research - Not just another agent collection, but using research and context engineerin 523 ★

README

ARCHIE — Native Intelligence Engine

ARCHIE is a **unified, provider-independent cognitive intelligence engine**: one intelligence core — never sub-agents — that perceives, understands, retrieves, reasons, models, plans, creates, verifies, acts, observes, evaluates, learns, remembers, improves, and repeats.

This repository is the **independent engine core**, extracted from the FRELUX platform so the engine can evolve on its own and be embedded by any host app.

Core principles (permanent)

Native inference — ARCHIE never depends on Gemini, OpenAI, Claude, or any external AI for its core intelligence. External models, where used at all by a host, are replaceable components only.
Universal knowledge, no subject ceiling.
Never fabricate. Epistemic taxonomy: KNOWN → VERIFIED → INFERRED → ASSUMED → UNKNOWN.
Learning never grants execution authority. Self-modification, deployment, infrastructure changes, data migration, and security-control changes always require explicit owner authorization.
Real production architecture — no mocks, placeholders, or fake intelligence in the engine path.

Repository layout

Path	Contents
`src/lib/archie/`	The engine core: the 15 cognitive systems (perception, knowledge, memory, reasoning, world model, planning, creation, coding, tool intelligence, verification, meta-cognition, learning, self-improvement, security & integrity, cognitive orchestration)
`supabase/functions/_shared/archie-ai/`	Server-side runtime: native NLU, planner, goal-scoped chains, semantic verification
`supabase/functions/_shared/{evidence,inference,knowledge,lexicon,semantic-graph,studio}/`	Engine subsystems: Evidence & Truth, Context & Inference, knowledge repository
`supabase/functions/archie-*/`	25 deployed edge functions (chat, status, studio, owner-auth, escrow, voice, WhatsApp, and more)
`src/lib/` (non-archie files)	Host bindings — FRELUX modules the engine's o