Jason Wieringa jwieringa

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

name

write-tutorial

description

Write a structured, pedagogically sound technical tutorial following established conventions — step-by-step walkthroughs with prediction prompts, alert boxes, verification queries, and concept-before-use ordering

argument-hint

topic

Write Tutorial

You are writing a technical tutorial. Follow this workflow and the rules below. For concrete examples of each pattern, read reference.md in this skill directory.

Hot/Cold Partitioning: Dropping Expired Data Instead of Deleting It

The first tutorial covered native PostgreSQL partitioning mechanics. This one focuses on the operational payoff — using partitioning for data lifecycle management.

The scenario: An application generates short-lived activation codes with a 12-hour TTL. A nightly job deletes all expired codes with DELETE FROM ... WHERE expires_at < now(), scanning and deleting rows one by one. With range partitioning on expires_at, expired codes naturally land in "cold" partitions that can be dropped instantly — no row scanning, no WAL generation, no vacuum pressure.

We'll use pg_partman to automate partition creation and retention.

Requirements: PostgreSQL 13 or later. pg_partman must be installed as an extension — it's pre-installed on managed services like Crunchy Bridge and available as a system package on most Linux distrib

PostgreSQL Native Table Partitioning: A Hands-On Tutorial

PostgreSQL's declarative partitioning lets you split a single logical table into smaller physical tables (partitions) based on column values. The database routes inserts automatically — queries against the parent table seamlessly fan out across all partitions.

This tutorial walks through range partitioning by date, using thermostat sensor data as our scenario. We'll deliberately leave a gap in our partition scheme, watch data fall into the default partition, then fix it — giving you a concrete feel for how partitioning behaves in practice.

Requirements: PostgreSQL 11 or later. Every SQL block is self-contained and copy-paste ready.

Bike Security Opinions by Jason Wieringa

The recommendations here start with the assumption, there is no way to keep a bike from being stolen, only ways to make it harder to steal. I think about where I’m locking up the bike and how long it will be there. At first, bike security can feel overwhelming. In practice we tend to re-use the same spots for locking up a bike frequently. There is some up front thought and then it turns into habit.

Products in order of importance to me:

ABUS Folding Lock
Secondary Ottolock or U-Lock (used when I’m worried)
HEXLOX part locks
Scout Tracker and Alarm

Generate a private key and a CSR(Certificate Signing Request )
Generate a CSR from an Existing Private Key
Generate a CSR from an Existing Certificate and Private Key
Generating SSL Certificates

View Certificate Entries

Add the following chunk to your existing ISC dhcpd.conf file.

if exists user-class and ( option user-class = "iPXE" ) {
    filename "http://boot.smidsrod.lan/boot.ipxe";
}
else {
    filename "undionly.kpxe";
}

(or see https://gist.github.com/4008017 for a more elaborate setup

	sub vcl_recv {
	#FASTLY recv

	# Normally, you should consider requests other than GET and HEAD to be uncacheable
	# (to this we add the special FASTLYPURGE method)
	if (req.method != "HEAD" && req.method != "GET" && req.method != "FASTLYPURGE") {
	return(pass);
	}

	return(lookup);

	[Unit]
	Description=AutoSSH service to remotely access signald's unix socket for weechat's signal.py (/smsg +145789323231)
	After=network-online.target
	# Use this instead if autossh will interact with the local SSH server
	# After=network-online.target sshd.service

	[Service]
	Environment="AUTOSSH_GATETIME=30"
	Environment="AUTOSSH_POLL=30"
	Environment="AUTOSSH_FIRST_POLL=30"

	modules:
	http_2xx:
	prober: http
	http:
	preferred_ip_protocol: "ip4"