Discover gists

NaiveProxy: Ультимативный гайд по настройке (2026)

Инструкции по установке

1. Подключаемся к серверу:

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.

[Your Name]'s AI Workspace

One flat folder. One memory file. One index that points your AI at what it needs.

How This Works

This file is the single entry point for your AI assistant. Instead of 15 folders with custom configs (ask me how I know that doesn't work), everything routes through here. Your AI reads this file, knows where to look, and gets to work.

Workspace Structure

Let's Encrypt SSL with Cloudflare DNS Validation

Requirements Recap

✅ Your Requirements:

🔒 Ports 80/443 are firewalled to Cloudflare IPs only
☁️ You're behind Cloudflare, so Let's Encrypt (Certbot) cannot reach port 80 directly
🌐 You're using Apache, not Nginx
You already have a self-signed or manually managed cert in use

	# Courtesy of: https://stackoverflow.com/a/11974399

	{%- for item in items %}
	[
	"{{item}}"{{ "," if not loop.last }}
	]
	{%- endfor %}

	#!/usr/bin/env python3
	"""Claude Code status line: API-rate cost estimate (session / today)."""
	import json
	import sys
	import glob
	import os
	import datetime as dt

	# USD per 1M tokens: (input, output, cache_write_5m, cache_read)
	# Source: platform.claude.com/docs/en/about-claude/pricing

	-- Step 1: Build daily event time series (last 90 days, excluding yesterday)
	WITH all_events AS (
	SELECT
	PARSE_DATE('%Y%m%d', event_date) AS time_series_data,
	event_name,
	COUNT(*) AS total_events
	FROM
	`your_project.analytics_XXXXXXXXX.events_*`
	WHERE
	_TABLE_SUFFIX BETWEEN

	#!/usr/bin/env python3
	"""
	Patch the installed Codex macOS app by editing its Electron ASAR webview bundle.

	WARNING: Modifying files inside `/Applications/Codex.app` will break the app's code signature.
	You may need to re-sign the app (or adjust Gatekeeper settings) after patching.

	Codex.app also enables Electron's ASAR integrity check. After repacking `app.asar`, you must
	update `ElectronAsarIntegrity` in `Codex.app/Contents/Info.plist`, otherwise the app will exit
	on startup with: