angusdev

LLM Skills Site – Design & Implementation Specification

1. Objective

Reimplement a local, static version of a skills directory (similar to skills.sh) using:

• React
• Vite
• Static JSON + Markdown data
• No backend

angusdev

You are generating documentation for an AI agent (not a human developer).

The goal is to produce a spec that an automated agent can use to:

• understand the purpose of the component
• know exactly what inputs it can send
• know exactly what outputs and errors it may receive
• know what the component does NOT do
• know any constraints, invariants, or assumptions

Requirements:

angusdev

sequenceDiagram
    participant Client
    participant API
    participant Blob as Blob (XYZ Data)
    participant DB as DB
    participant Worker as Background Worker (AKS)
    participant Cache as Blob (SQL Cache)
    participant SourceDB as Source Database

    Client->>API: Request (XYZ, fromDate, toData)

angusdev

SplitBrowser

SplitBrowser is a resizable, multi-column mini browser bookmarklet that lets you open and navigate multiple websites side-by-side in a single tab. Each column has independent navigation and history, and URLs are tracked live for easy sharing. Ideal for sharing in Teams, comparing sites, monitoring dashboards, or running parallel tests.

⚠️ This project and README are 100% AI-GENERATED from spec.md and have not been modified by any human.
Developed using a Spec-Driven Development approach.

---

Features

angusdev

import uuid
from abc import ABC, abstractmethod
from typing import Any, Optional

import dash
import flask
from dash import Input, Output, State, dcc, html
from dash.development.base_component import Component

angusdev

change_favicon_color=(c)=>{u=window.top.document.querySelector('link[rel="icon"], link[rel="shortcut icon"]');if(u.href.startsWith('data:')){console.log('Favicon is already a data URL.');return;}fetch(u.href).then(r=>r.blob()).then(b=>{i=new Image();i.src=URL.createObjectURL(b);i.onload=()=>{a=document.createElement('canvas'),ctx=a.getContext('2d');a.width=i.width;a.height=i.height;ctx.drawImage(i,0,0);ctx.globalCompositeOperation='source-in';ctx.fillStyle=c;ctx.fillRect(0,0,a.width,a.height);u.href=a.toDataURL('image/x-icon');}})};

angusdev

import re
from datetime import datetime
import pandas as pd
import argparse

def parse_log_files(file_names):
    job_data = {}
    job_pattern = re.compile(r'(\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}) (Thread-\d+) MYJOB (started|ended) (\d+) (.+)')

    for file_name in file_names:

angusdev

from pyspark.sql import SparkSession
from pyspark.sql import Row

# Initialize Spark Session
spark = SparkSession.builder.appName("Vertical Display").getOrCreate()

# Example DataFrame
data = [("Alice", 25), ("Bob", 30)]
columns = ["Name", "Age"]

angusdev

from typing import List, Dict, Any, Tuple, Optional
from sqlalchemy import create_engine, text
from sqlalchemy.engine import Engine, ResultProxy
import pandas as pd
from pyspark.sql import SparkSession, DataFrame as SparkDataFrame

# Base class DataAccess using SQLAlchemy
class DataAccess:
    def __init__(self, conn_str: str):
        self.engine: Engine = create_engine(conn_str)

angusdev

# apple_runner.py

class AppleRunner:
    def run(self):
        print("Apple")