Build a full-stack project from scratch in Copilot agent mode
The poster built a full-stack project with Copilot agent mode using Claude 3.7 Sonnet; it needed babysitting through setup problems but eventually produced styled, working React components.
Done withGitHub CopilotClaude
- What they did
- The poster used GitHub Copilot agent mode to build a full-stack app from scratch across React, TypeScript, Tailwind, ShadCN/UI, Python, Clojure, and PostgreSQL. Before each run they gave the agent three files: a project spec, a step-by-step task list, and a Copilot instructions file; then they babysat shell commands, corrected bad paths and Windows/Linux CLI mismatches, and kept refining the instructions between runs.
- How it went
- Claude 3.7 Sonnet was the most proactive and kept trying to recover from setup errors, but all models initially failed on the React, Tailwind v4, and ShadCN setup. After a second Claude run and better instructions, the poster says they finally got fully styled, working React components.
- Worth knowing
- The main catch was setup reliability: the poster says even routine install and directory steps needed repeated manual intervention and instruction-file tweaks.
Try it yourself with GitHub Copilot
Using agent mode, build a full-stack project from scratch: [describe the app] with a [frontend, e.g. React] frontend and a [backend/database] backend. Set up the project structure, get it running locally, and create styled components for [main screens or features]. If setup hits an error, try to fix it yourself first, and stop to ask me before installing anything large or changing config outside the project folder.
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: Reddit · Undated
Related use cases
Build a C compiler with a team of parallel agents
Anthropic's engineering write-up on tasking Opus 4.6 agent teams with building a C compiler.
Generate a daily personalized trading research report from brokerage data
Lonnie VanZandt's Claude-based assistant reads his Schwab positions and watchlist via MCP, sweeps news and Fed sources, and writes a daily report.
Take orders for a sushi chain with an agent
The poster built a Claude agent that takes orders for a seven-location sushi chain, using menu, allergen, delivery zone, hours and promo data.
Turn tickets into merged pull requests
Pulls tasks from GitHub Issues or Linear, then runs Claude Code or Codex in isolated Kubernetes pods to carry each one through to a PR.
Fix a broken Delete button in a GitHub repo
The author had Jules clone a repo and fix a Delete button so it disables deletion of content older than two hours, with tests added unprompted; the author reviewed and published the branch.
Run a multi-bot team that opens draft PRs
A Grok setup of research, classifier, builder, tester, reviewer, and security bots maps repos, makes minimal changes, runs tests and lint, and opens draft PRs; the author reports over 1,000 PRs.
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.