{
  "id": 20913605,
  "title": "I gave Claude Code a two-hour billing project. CLAW ran the team.",
  "url": "https://urgent.news/2026/11/20/i-gave-claude-code-a-two-hour-billing-project-claw-ran-the-team",
  "topic": "ai",
  "section": "AI",
  "published": "2026-11-20T15:09:06.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/290990collab/i-gave-claude-code-a-two-hour-billing-project-claw-ran-the-team-4gh6"
  },
  "original_language": "en",
  "account": "When Claude Code is given clear instructions, a method for delegation, and knowledge on which model to utilize for each task, remarkable results can be achieved. The reporter decided to test this theory by giving Claude Code a complex billing service project, complete with subscriptions, usage metering, invoices, webhooks, dunning, credit notes, a double-entry ledger, and more. Additionally, the reporter asked Claude Code to intentionally introduce 16 bugs, each with a hidden test to identify them. In just two hours, Claude Code produced a 22,000-line project, half of it comprised of tests, and successfully identified all 16 seeded bugs using their hidden tests. The project was completed by seven Sonnet agents, utilizing 91% of tokens, while Opus coordinated the work, accounting for 9% of tokens. The reporter found that CLAW, a system designed to give Claude Code structure and consistency, played a crucial role in the project's success. CLAW provides Claude Code with a shared working method, a team of specialized agents for different tasks, and project memory to keep track of decisions and progress. This approach proved effective not only in building the billing service but also in navigating and fixing the deliberately introduced bugs.",
  "summary": "What happens when you give Claude Code more than just a prompt? When it has a clear way of working, knows when to delegate, understands which model to use for each job, and has rules for checking its own work? I wanted to find out. So I gave it a fairly ambitious task: build a B2B billing service, close enough to production to be a meaningful test. Subscriptions, usage metering, invoices with VAT…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}