Built with Claude Code Study Portal

TL;DR

This platform was designed, built, and is maintained largely with Claude Code — an agentic AI coding tool that can read and write files, run shell commands, search the codebase, and orchestrate sub-agents. The workflow is not "ask an AI to write some code and paste it in." It is a structured system of specialized subagents, a mandatory verification pipeline, parallel delegation across domains, and a feedback loop that runs all checks through real tooling. The human owner sets direction, makes the architecture calls, and gates anything irreversible. The AI accelerates implementation, review, and testing inside tight, automated verification loops — the value comes from those loops, not from blind code generation.

If you remember one thing: the project does not trust AI claims. Every change is verified by actually running the lint, tests, build, and end-to-end suites and reading the real results.