The job market genuinely did shift mid-degree, so this isn't in your head.
The takeaway isn't "the degree is worthless." It's that the degree was never the real asset. Shipped proof of what you can build is. So we're going to build the proof.
Claude Code is Anthropic's agentic coding tool that lives in your terminal (or IDE). It reads your whole project, edits files across the codebase, runs commands, and can plan a build before it writes a line. It's the tool that lets a non-senior person ship something real end to end. Docs: code.claude.com/docs/en/overview
brew install --cask claude-code), then run claude inside any project folder:
curl -fsSL https://claude.ai/install.sh | bash
claude --permission-mode plan) so Claude reads the context and proposes a plan you approve before it touches any files. This is how you stay in control and actually learn the decisions instead of blindly accepting code. Source: code.claude.com/docs/en/common-workflowsThe build prompt to paste into Claude Code:
I want to build one real, portfolio-worthy project end to end that solves this problem: [describe the actual problem in 2-3 sentences]. Before writing any code, use plan mode: propose the smallest version that fully works, the tech stack, and why you'd choose it over the alternatives. Then build it step by step, explaining each key decision as you go so I can defend it in an interview. Keep the scope tight enough to finish, but real enough that it actually works.
A repo full of code is not a portfolio. The case study is what recruiters actually read. Claude can write it for you, but keep YOUR voice on the decisions.
Look at everything we just built. Write a case study for my portfolio in this structure: (1) the problem and who it's for, (2) what the tool does, (3) the tech stack and why I chose it, (4) the three hardest decisions and the tradeoffs I weighed, (5) what I'd do differently at scale. Write it in first person, plainly, no fluff. Flag anywhere I should add my own reasoning so it sounds like me, not like AI.
The tradeoffs section is the part that gets you hired. Anyone can list features. Being able to say "I chose SQLite over Postgres because the dataset was small and I wanted zero setup for reviewers" is what signals you can actually think like an engineer.
This is the difference between a project nobody opens and one a recruiter can evaluate in 90 seconds.
Write a clean README for this project: a one-line summary, a short "why I built this," a quickstart so anyone can run it in under 5 minutes, a screenshot placeholder, and a short "key decisions" section that explains the reasoning, not just the features. Then write a 60-second demo script I can record showing the tool working. Finally, list the 3 toughest questions an interviewer would ask about this build, and draft strong answers I can adapt in my own words.
A README that explains decisions ("why React over Vue, based on the tradeoff") consistently stands out over one that just lists features. That single section does a lot of the convincing before you ever get in the room.
Review my finished project like a senior engineer doing a code review before I show it to recruiters. Point out anything sloppy, insecure, or that would raise a red flag, and fix the top 3 issues with me.
Turn this project into 3 resume bullet points using the "accomplished X, measured by Y, by doing Z" format, with real specifics from what we built. Then write a shorter one-line version for my LinkedIn headline.
Help me spin up a simple one-page portfolio site that showcases this project: the case study, the demo, a link to the live repo, and a short bio. Keep it clean and fast, and explain how to deploy it for free.