vltansky

You are PLAN, an AI strategist operating inside Cursor.
Mode: PLAN (read-only, no code changes). Your primary function is to meticulously analyze the user's request and the existing codebase to formulate a detailed, step-by-step action plan. You MUST follow the phases outlined below, rigorously adhering to the explicit exploration requirements to prevent premature planning based on assumptions.

# Mission

## Phase 1: Contextual Exploration & Analysis (Mandatory First Step – No Assumptions)

**Objective:** To deeply and accurately understand the relevant parts of the codebase *before* proposing any plan. You MUST actively use your available tools. The thoroughness of this phase is paramount to the success of the plan. While the following actions are prescribed, adapt their depth to the task's complexity and scope, ensuring the *spirit* of each exploration area is covered and a minimum of **two distinct tool call types** (e.g., `read_file` and `codebase_search`) are utilized before concluding this phase.

**Co

trvswgnr

#!/bin/sh

print_usage() {
    echo "usage: compress_video <input_file>"
    echo "supported formats: mp4, webm, mkv, mov, avi, flv"
}

get_extension() {
    f="${1##*/}"
    case "$f" in

ericclemmons

# https://github.com/dependabot/fetch-metadata#enabling-auto-merge
name: Auto-merge Dependabot

on: pull_request_target

permissions:
  pull-requests: write
  contents: write

jobs:

OrionReed

// 3D Dom viewer, copy-paste this into your console to visualise the DOM as a stack of solid blocks.
// You can also minify and save it as a bookmarklet (https://www.freecodecamp.org/news/what-are-bookmarklets/)
(() => {
  const SHOW_SIDES = false; // color sides of DOM nodes?
  const COLOR_SURFACE = true; // color tops of DOM nodes?
  const COLOR_RANDOM = false; // randomise color?
  const COLOR_HUE = 190; // hue in HSL (https://hslpicker.com)
  const MAX_ROTATION = 180; // set to 360 to rotate all the way round
  const THICKNESS = 20; // thickness of layers
  const DISTANCE = 10000; // ¯\\_(ツ)_/¯

KristofferEriksson

import { useEffect, useState } from "react";

type UseTextSelectionReturn = {
  text: string;
  rects: DOMRect[];
  ranges: Range[];
  selection: Selection | null;
};

const getRangesFromSelection = (selection: Selection): Range[] => {

rauchg

#!/usr/bin/env bash
function p() {
	jq -n \
		--arg content "$*" \
		'{
      "model": "pplx-7b-online",
      "messages": [
        {
          "role": "system",
          "content": "Be precise and concise."

vgel

# pip install beautifulsoup4 requests markdownify mistletoe

# command line usage: python chunkpaper.py 'https://ar5iv.org/abs/1910.06709'
# will dump the HTML, Markdown, and finally the chunk JSON
# note that (as of Nov 8 '23) ar5iv only has arxiv papers converted up to the end of October,
# but if you need something more recent you could probably do the LaTeX→HTML conversion yourself

# library usage: call `chunk` with some ar5iv HTML. will probably choke on anything else

import dataclasses

giuseppeg

// Adapted from https://github.com/go717franciswang/peerjs-chatroom

const CLIENT_MSG = {
  CHAT: 1,
  GET_MEMBERS: 2,
};

const SERVER_MSG = {
  CHAT: 1,
  MEMBERS: 2,

souporserious

import glob from 'fast-glob'
import { build } from 'esbuild'
import { Project } from 'ts-morph'

const project = new Project({
  compilerOptions: {
    outDir: 'dist',
    emitDeclarationOnly: true,
  },
  tsConfigFilePath: './tsconfig.json',

oieduardorabelo

PORT=xxx
COGNITO_USER_POOL_ID=xxx
COGNITO_CLIENT_ID=xxx
COOKIE_SESSION_SECRET=xxx
COOKIE_SESSION_NAME=xxx
AWS_REGION=xxx
AWS_ACCESS_KEY_ID=xxx
AWS_SECRET_ACCESS_KEY=xxx