tommyjtl

Minimal coding agent on Modular Cloud

A small coding agent that builds a single-page web app from one prompt, using a model on Modular Cloud's shared endpoints through its OpenAI-compatible API. It's one loop: call the model, run any tool calls it asks for (write_file, read_file), send the results back, and stop when it's done.

Run it

# optionally, set up your virtual environment first
pip install openai
export MODULAR_API_KEY=your-key-here

tommyjtl

The standard Lorem Ipsum passage, used since the 1500s
"Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum."

Section 1.10.32 of "de Finibus Bonorum et Malorum", written by Cicero in 45 BC
"Sed ut perspiciatis unde omnis iste natus error sit voluptatem accusantium doloremque laudantium, totam rem aperiam, eaque ipsa quae ab illo inventore veritatis et quasi architecto beatae vitae dicta sunt explicabo. Nemo enim ipsam voluptatem quia voluptas sit aspernatur aut odit aut fugit, sed quia consequuntur magni dolores eos qui ratione voluptatem sequi nesciunt. Neque porro quisquam est, qui dolorem ipsum quia dolor sit amet, cons

tommyjtl

tommyjtl

pdf-compress() {
    QUALITY=150
    INPUT_PDF=""
    GS_BIN=$(command -v gs)

    print_usage() {
        echo "Usage: pdf-compress [-q|--quality <res>] <file>"
        echo "    Where:"
        echo "        -q|--quality <res> is the resolution for the PDF (default: 150)"
        echo "        <file> is the PDF file to compress"

tommyjtl

enter() {
    local config_dir="$HOME/.config/enter"
    local config_file="$config_dir/enter.yml"

    # Initialize configuration
    if [ ! -d "$config_dir" ]; then
        mkdir -p "$config_dir"
    fi
    if [ ! -f "$config_file" ]; then
        local user_root=$(pwd )

tommyjtl

python-arm() {
    # Get the platform information
    platform=$(uname -m)

    # Check if the platform is darwin arm
    if [[ "$platform" == "arm64" ]]; then
        # Run the desired command if on darwin arm platform
        /opt/homebrew/opt/python@3.9/bin/python3.9 "$@"
    else
        # Exit if not on darwin arm platform

tommyjtl

This gist demonstrated the example use of serving secured WebSocket server (wss://) using Python's websockets via Nginx and PM2 the process manager. Have not tested thoroughly on production, so this is only for development use.

• For fullchain.pem and privkey.pem, you will need use cerbot to generate them.
• For pm2.json, run pm2 start pm2.json to run the WebSocket server in the background.

tommyjtl

function cocomod() {
    # port          list all detected ports:                        python3 -m there detect
    # run           run local file on module (not send):            python3 -m there run [FILE]
    # list          list file:                                      python3 -m there ls -l [/]      python3 -m there run [FILE]
    # send          send file to module:                            python3 -m there push [FILE] [/]
    # read          get source code from file on module:            python3 -m there cat [FILE] 
    # delete        delete file from module:                        python3 -m there rm [FILE]
    # repl          enter repl:                                     python3 -m there -i
    # echo "Usage: burn, goto, port, run, list, send, read, delete, repl" 

tommyjtl

import subprocess, os

current_path = os.getcwd()
current_file_list = os.listdir()

def exec_command(cmd):
    command = cmd.split()
    command_output = subprocess.run(command, capture_output=True)
    for line in command_output.stdout.decode("utf-8").split("\n"):
        print("\t\t", line)

tommyjtl