Build Log
What is vibe coding, honestly? A builder who ships real products with Cursor and Claude Code on where the AI wins, where it breaks things, and who it's for.
A Cursor AI tutorial that takes one failing test to a reviewed commit through Plan Mode, the Run Mode that decides whether tests run unasked, and Agent Review.
Multi agent AI helps for three reasons and costs near 15x the tokens. A test for each, a context calculation, and the smallest two-agent setup worth building.
The five types of AI agent from the textbook that defined them, each as ten lines of Python you can run, plus which type an LLM tool-calling loop actually is.
How to build an AI agent in one Python file: three tools, a loop with a turn cap and a token cap, a run log, and a check that grades the run from data.
One app spec held against the AI app maker tools on page one, with each free tier's documented limits, the walls you hit, and the money outside the maker.
What an autonomous AI agent can do without a human, mapped to the permission modes, sandboxes and hard caps that Claude Code, Codex and Copilot document.
A one-feature spec, a six-row rubric, and a mode matrix for judging any AI coding assistant yourself, plus today's free tiers and prices from vendor pages.
What an MCP server is, a minimal Python server you can run, the Claude Code command to register it, and the failure modes that make a first server go silent.
Test AI-generated code across contracts, boundaries, permissions, migrations, retries, concurrency, and rollback with a reusable failure matrix.
Use an AI coding workflow that turns issue contracts, tests, diffs, review findings, and release checks into evidence instead of confident summaries.
Use an AGENTS.md file to give coding agents scoped repository rules, nested overrides, exact verification commands, and deployment anti-drift boundaries.
Follow a Claude Code tutorial that inspects, plans, edits, and verifies one small repository change with exact checkpoints and test evidence.
Compare Codex vs Claude Code with the same five repository tasks, evidence checklist, permission boundaries, and a decision method you can reproduce.
The real vibe coding security risks a shipping dev keeps hitting, from hardcoded secrets to missing auth checks, plus the short guardrail list that stops them.