ml-eds

Designing a Personal-Guidance LLM Agent: A Meta-Guide

Purpose: A reusable design framework for building LLM agents that give personal guidance in any high-stakes domain — mental health, career, financial, parenting, relationships, learning. Sections 1–6 are domain-agnostic instructions. Section 7 is a worked case study (mental health therapy) showing every rule instantiated. Last updated: 2026-05-04 ⚠️ Disclaimer: AI is not a substitute for licensed professionals. Any agent built from this guide must include crisis escalation and professional referral paths.

---

1. When to use this guide

ml-eds

The Mental Model for Success: From Software Engineer to Life

A Lattice, Not a Single Idea

Charlie Munger's key insight was that no single mental model is enough. You need a latticework — a connected web of thinking tools from many disciplines. The models hang together and reinforce each other. The person with 80–90 well-understood models across psychology, engineering, economics, and biology will consistently outperform the person who only thinks in code.

Here's a structured lattice, from craft to life:

---

ml-eds

SSL for Postgres connections

In the following we will create a self-signed certificate (without self-signed Root CA certificate, to keep things simple)

postgresql.conf

In /var/lib/postgresql/data:

nano postgresql.conf

ml-eds

# export all changed files from commit b58b5f9 to commit HEAD into tar archive

git diff-tree -r --no-commit-id --name-only --diff-filter=ACMRT b58b5f9^..HEAD | tar -czf file.tgz -T -

ml-eds

# Convert Unix timestamp to date in zsh/bash
# 
# Add to .zshrc or .bash_profile 
# 
# Example usage:  ts2date 1650621128
# Example output: 2022-04-22 11:52:08

timestamp2date() {
     date -r "$1" '+%Y-%m-%d %H:%M:%S' 
}

ml-eds

#!/bin/bash

# given list_search.txt with content, e.g.
# <PackageReference Include="Microsoft.EntityFrameworkCore.Design" Version="3.1.0" />
# <PackageReference Include="Microsoft.EntityFrameworkCore.InMemory" Version="3.1.0" />

# given list_replace.txt with content, e.g.
# <PackageReference Include="Microsoft.EntityFrameworkCore.Design" Version="3.1.5" />
# <PackageReference Include="Microsoft.EntityFrameworkCore.InMemory" Version="3.1.5" />

ml-eds

# good reference: https://wiki.ubuntuusers.de/CA/

# download missing cert DigiCertHighAssuranceEVRootCA.crt 
wget --no-check-certificate https://dl.cacerts.digicert.com/DigiCertHighAssuranceEVRootCA.crt

# convert to pem format
openssl x509 -inform DER -outform PEM -in DigiCertHighAssuranceEVRootCA.crt -out DigiCertHighAssuranceEVRootCA.pem.crt

# copy to /usr/local/share/ca-certificates
cp DigiCertHighAssuranceEVRootCA.pem.crt /usr/local/share/ca-certificates/

ml-eds

// Place the following in the "Tests" tab of a request

// Check HTTP status code 
tests['Status 200'] = responseCode.code === 200;

// We need valid JSON
var validJSON = false;
var resData = {};
try {
    resData = JSON.parse(responseBody);

ml-eds

TSLint pre-commit hook

Important!

Must be configured on top level directory of a project when having a multi-package project (e.g. frontend + backend).

Install

npm install --save-dev husky lint-staged

ml-eds

find . -type f -newermt '7/14/2019 16:00:00'