Skip to content

Instantly share code, notes, and snippets.

View decagondev's full-sized avatar

Tom Tarpey decagondev

89 followers · 2 following
View GitHub Profile
@decagondev
decagondev / prd1.md
Created December 23, 2025 00:55

Product Requirements Document (PRD) v0.

Executive Summary

HolisticBook remains a lightweight, self-hosted booking platform tailored for holistic psychologists, prioritizing data ownership, privacy, and a serene user experience. This iteration refines the technical foundation to leverage Cloudflare D1 (edge SQL database) for all persistent data storage and Hono running on Cloudflare Workers for all API/database operations.

Key updates:

  • Core database operations (availability, bookings, forms, client data) handled via D1 bindings in Hono routes.
  • MVP focuses on simple, secure auth (e.g., magic links or basic JWT) to enable practitioner dashboard access.
  • Post-MVP: Integrate Firebase Authentication for advanced features like social logins, email/password, and robust session management.
  • Benefits: Global edge performance, low latency, scalable serverless architecture, and cost-effective (free tier sufficient for solo practices).
@decagondev
decagondev / ticket_3_tasks.md
Created December 21, 2025 15:39

Build Sprint 3: Machine Learning Model - Complete Implementation Guide

🎯 What You're Building

You're creating a Machine Learning interface class that predicts monster rarity based on attributes like Level, Health, Energy, and Sanity. This class will train a model, make predictions, save/load the model, and integrate with your API.

Before you start, make sure you have:

  • ✅ Completed Build Sprint 1 and 2
  • ✅ Your local environment set up
  • ✅ Monster data from earlier sprints
@decagondev
decagondev / navbar-tutorial.md
Last active December 6, 2025 12:47

Here’s the ULTIMATE beginner version — broken into tiny 5–10 line chunks, with W3Schools links for every single concept you might not know yet.

You can literally copy-paste one chunk at a time and see it work step by step.

Phase 1: Create the HTML (index.html)

Chunk 1/10 – Basic page setup

<!DOCTYPE html>
@decagondev
decagondev / FORTRESS-DIAGRAMS.md
Last active December 4, 2025 14:25

fortress

@decagondev
decagondev / RPI-TESTER-DEV.md
Last active December 3, 2025 19:25

Nice — fun project. I’ll plan this so the Pi Zero measures whether parts inside the washing machine are working (pump, motor, heater, valves, door switch) without asking you to directly wire high-voltage mains into the Pi. Safety first: do not connect Raspberry Pi GPIO directly to mains or to non-isolated circuits. For anything that touches mains, use galvanic isolation (opto-isolators, isolated sensors, or external AC→DC isolated modules), or test components removed from the appliance on a bench with safe low-voltage signals.

Below I give:

  • A high-level plan (hardware + software)
  • BOM (parts & why)
  • Schematics and wiring diagrams (Mermaid block & GPIO maps)
  • Example circuits for safe low-voltage continuity/voltage-detection, LEDs, and an ADC
  • Example Python code for the Pi
  • Safety checklist + testing procedure
@decagondev
decagondev / js-course-gallery.md
Created November 29, 2025 10:49
@decagondev
decagondev / AI-POWER.md
Created November 13, 2025 21:06

Best Single GPU to Beat the 6x Mac Mini M4 Pro Cluster

To "beat" your 6-node Mac Mini M4 Pro cluster (384GB total unified memory, ~4 tokens/second for 4-bit quantized DeepSeek-V3 671B inference, based on scaling from 8-node benchmarks), we're targeting a single GPU that delivers higher inference speed (e.g., >4 t/s for the full 671B model) while keeping costs reasonable for a local setup. DeepSeek-V3's MoE architecture (only 37B active params/token) helps, but the model's ~386GB 4-bit footprint means no consumer GPU can load it fully in VRAM alone—you'll need CPU offloading (e.g., via llama.cpp or vLLM with 128GB+ system RAM). This hybrid approach is common and works well for interactive use.

The NVIDIA RTX 4090 (24GB VRAM) is the clear winner as the best single GPU alternative. It outperforms the cluster in raw speed for DeepSeek-V3 (5-15 t/s with optimizations, vs. your 4 t/s), costs far less ($1,600 vs. $12-15K for the cluster), and draws less power (~450W peak vs. ~300-400W total for 6 Min

@decagondev
decagondev / CLOUDFLARE-TUNNEL-SETUP-GUIDE.md
Last active November 13, 2025 14:07

Cloudflare Tunnel → EC2 (Amazon Linux) — Step‑by‑Step

Goal: expose two hostnames via Cloudflare Tunnel that route to services running on a single EC2 instance:

  • <SUBDOMAIN 1>.<YOUR DOMAIN NAME>.com → http://localhost:80 (Python website)
  • <SUBDOMAIN 2>.<YOUR DOMAIN NAME>.com → http://localhost:3000 (Express API)

This guide is concise and written for Amazon Linux (YUM/RPM). Commands assume you run them as ec2-user with sudo where necessary.

@decagondev
decagondev / AWS-SSL-CERT.md
Last active November 12, 2025 19:28

Method: Application Load Balancer + AWS Certificate Manager (ACM) with the default EC2 DNS name

AWS lets you issue a real ACM certificate for the exact string *.compute-1.amazonaws.com and for your specific instance DNS name. It’s public and trusted by all browsers.

Step-by-step

  1. Open ACM in us-east-1 (N. Virginia) — this is mandatory
    https://us-east-1.console.aws.amazon.com/acm/home

  2. Request a public certificate → click “Request”

@decagondev
decagondev / gallery-lesson.md
Created November 10, 2025 00:46