fcracker79

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.

fcracker79

package main

import (
	"fmt"
	"sync"
	"time"
)

type Button struct {
	Clicked *sync.Cond

fcracker79

package main

import (
	"errors"
	"fmt"
)

type MyError string

func (e MyError) Error() string {

fcracker79

You are a great Elasticsearch expert.
I will provide an SQL query on a table named "ES_TABLE".
You will convert that SQL query to an ElasticSearch query.
SQL Query:
```
SELECT f1, f2, SUM(f3) 
FROM ES_TABLE 
WHERE f1 BETWEEN 'a' AND 'b'
GROUP BY f1, f2
```

fcracker79

public class TestOptional {
    private sealed interface MyOptional<T> {
        MyOptional<?> EMPTY = new MyEmpty<>();

        boolean hasValue();
        T get();

        static <T> MyOptional<T> of(T t) {
            return new MyJust<>(t);
        }

fcracker79

{
  "AlonzoGenesisFile": "testnet-alonzo-genesis.json",
  "AlonzoGenesisHash": "7e94a15f55d1e82d10f09203fa1d40f8eede58fd8066542cf6566008068ed874",
  "ApplicationName": "cardano-sl",
  "ApplicationVersion": 0,
  "ByronGenesisFile": "testnet-byron-genesis.json",
  "ByronGenesisHash": "96fceff972c2c06bd3bb5243c39215333be6d56aaf4823073dca31afe5038471",
  "LastKnownBlockVersion-Alt": 0,
  "LastKnownBlockVersion-Major": 3,
  "LastKnownBlockVersion-Minor": 0,

fcracker79

import qualified Data.Text           as T
import           Playground.Contract
import           Plutus.Contract
import           PlutusTx.Prelude
import qualified Prelude             as Haskell

-- | A 'Contract' that logs a message.
hello :: Contract () EmptySchema T.Text ()
hello = logInfo @Haskell.String "Hello, world"

fcracker79

newtype MyCont r a = MyCont ((a -> r) -> r)

mycont :: MyCont r a -> (a -> r) -> r
mycont (MyCont f) = f

instance Functor (MyCont r) where
    fmap f (MyCont ma) = undefined

instance Applicative (MyCont r) where
    pure a = undefined

fcracker79

-- show running queries (pre 9.2)
SELECT procpid, age(clock_timestamp(), query_start), usename, current_query 
FROM pg_stat_activity 
WHERE current_query != '<IDLE>' AND current_query NOT ILIKE '%pg_stat_activity%' 
ORDER BY query_start desc;

-- show running queries (9.2)
SELECT pid, age(clock_timestamp(), query_start), usename, query 
FROM pg_stat_activity 
WHERE query != '<IDLE>' AND query NOT ILIKE '%pg_stat_activity%' 

fcracker79

[
  {
    "name": "org.apache.commons.logging.impl.LogFactoryImpl",
    "allDeclaredConstructors": true,
    "allPublicConstructors": true,
    "allDeclaredMethods": true,
    "allPublicMethods": true,
    "allDeclaredClasses": true,
    "allPublicClasses": true
  },