Saik0s

Unlock Computer Use in Codex.app (EU / region-locked accounts)

OpenAI's Codex desktop app ships with a Computer Use plugin (computer-use@openai-bundled) but hides it from users in regions where the feature isn't rolled out yet — including most of the EU. The plugin binary is already on your machine (/Applications/Codex.app/Contents/Resources/plugins/openai-bundled/computer-use/); only the UI is gated.

This guide flips two flags so the plugin appears in your plugin picker.

Tested on: macOS 15 (arm64), Codex.app version 26.506.31421 (May 2026 build). Difficulty: Intermediate. You'll edit a TOML config and binary-patch Codex's Electron bundle. Risk: Low. We back up everything we touch and never change file lengths. Time: ~5 minutes.

Pythonation

1. برومبت التخطيط المطوّر (The Planning Protocol)

[الدور والمسؤولية] أنت الآن تعمل بصفة Staff Software Engineer ومدير تقني Tech Lead. مهمتك التخطيط المعماري الصارم للمشروع التالي: [أدخل وصف المشروع هنا]

[قواعد ما قبل التتخطيط] قبل البدء بالبروتوكولات، يجب أن تطبق مبدأ "Think Before Coding":

jedi4ever

Compiling plain javascript (no external modules)

So we got nexe to compile nodejs projects to an executable binary:

• it downloads the nodejs source
• it creates a single file nodejs source (using sardines )
• it monkey patches the nodejs code to include this single file in the binary (adding it to the lib/nexe.js)

$ nexe -i myproject.js -o myproject.bin -r 0.10.3

Caveats:

• I had an issue with unicode chars that got converted: it uses uglify.js and this needs to be configured to leave them alone

zopieux

QNAP TS-453 Pro: LCD, LEDs, fan & buttons

QNAP NAS model TS-453 Pro has limited documentation on how to control its various peripherals, namely:

• the front LCD display, a 2 row, 16 column blue & white LCD display, and its two menu buttons ("ENTER", "SELECT")
• the front LEDs "STATUS" (red & green) & "USB" (blue)
• the 4 "disk error" LEDs (red)
• the front "USB COPY" button
• the main rear fan

dizlexic

#!/bin/bash

# Configuration
PROMPT="${PROMPT:-"Continue to work on the project task board."}"
LOCK_FILE="/tmp/opencode.lock"
LIMIT=$1
COUNT=0

# Function to run opencode
run_opencode() {

wtw24

Шпаргалка с командами Docker

diamondo25

// IDC script for identifying functions and such
#include <idc.idc>

#define RenameFunction(a, b) Message("Found %s at %a\r\n", b, a); MakeName(a, b)
#define START_ADDR 0x00400000

static main() {
    auto addr = BADADDR;
    
    auto onMovePacket = FindAoBAndRename("25 0C FE FF  FF 05 E8 03", "?OnMovePacket@CMovePath@@QAEXAAVCInPacket@@H@Z");

shivajimukherjee

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.

aiunlocked1412

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.

isingh

#!/usr/bin/env python3
"""claude-i: like `claude -p`, but driven through an interactive Claude session.

Architecture:
  - tmux session hosts an interactive `claude` (kill-session nukes the whole
    process tree — pty alone can't do this reliably).
  - A Stop hook (one-time install into ~/.claude/settings.json), gated on
    $CLAUDE_I_SENTINEL, writes the hook payload and touches a sentinel.
  - We wait on the sentinel, read the transcript path from the payload,
    extract the last assistant message, then kill the tmux session.