Adam Teale adamteale

Autonomous Product Development with Paperclip + BMAD + Hermes

TL;DR: A one-sentence product description goes in. A fully-developed product — briefs, PRDs, UX designs, architecture, code, tests, and documentation — comes out. No human in the loop beyond the initial prompt and optional review gates.

🚀 The 2026 M5 Max Local-First AI Guide (v3 — Subscription Shield Edition)

Local Power + $20/mo Unlimited Cloud Fallback

This document is the updated setup for a high-performance, local AI development environment on an M5 Max (36GB RAM). It optimizes the stack for the M5's Neural Accelerators and integrates the Alibaba Model Studio $20/mo Coding Plan as a fixed-price "Financial Firewall."

What changed from v2:

M5 Optimization: Updated thread counts and memory alignment for the M5 architecture.

Parity Routing: Switched fast-coder to Qwen-3.5 to match the cloud subscription logic.

The Shield: LiteLLM is now configured with a $1.00 hard budget ceiling to prevent pay-as-you-go bill shock.

🚀 The 2026 M4 Max Local-First AI Guide (v2 — llama.cpp Edition)

Ditching Ollama: Direct llama.cpp for 2.4x Faster Inference

This document is the updated setup for a high-performance, local AI development environment on an M4 Max (36GB RAM). It replaces the Ollama backend with llama-server (upstream llama.cpp), delivering a 2.4x speedup on the fast-coder slot and 1.9x on the think-brain slot, with correct model templating and full control over every flag.

What changed from v1: Ollama is gone as the inference backend. llama-swap stays as the proxy but now talks directly to llama-server instead of spawning Ollama processes. Every flag you set now actually applies.

🚀 The 2026 M4 Max Local-First AI Guide

Architecting the 110+ TPS "Senior-Junior" Local Brain

This document provides the definitive setup for a high-performance, local AI development environment on an M4 Max (36GB RAM). It leverages Speculative Decoding, AST-based Indexing, and Kernel-level Memory Tuning.

🏗️ 1. The Core Architecture (Senior-Junior)

On an M4 Max with 36GB, the "Sweet Spot" is running a 35B-parameter model (The Architect) and a 1.5B-parameter model (The Draft Model).

	version: "3.7"

	services:
	nutshell:
	image: cashubtc/nutshell:latest # The Docker image for Nutshell
	ports:
	- "3338:3338" # Expose port 3338
	environment:
	- MINT_BACKEND_BOLT11_SAT=FakeWallet
	- MINT_PRIVATE_KEY=TEST_PRIVATE_KEY



	<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
	<html xmlns="http://www.w3.org/1999/xhtml" lang="en-us" >
	<head>
	<title>Select asset to change \| MBrain admin</title>
	<meta name="viewport" content="width=device-width, initial-scale=1.0"/>