Projects & Explorations

Motivation: I wanted one small, MIT-licensed repo where I could master Cursor agent setup through vibe coding and experiments—not just read about rules and hooks, but wire them, break them, and see how they compose.

Problem: Most agent tutorials stop at prompts. In real work you need repo-native configuration: baselines that stick (rules), named procedures (skills), focused delegation (subagents), lightweight project intent (AGENTS.md), and automation around edits and shell (hooks). Without a map and working samples, you reinvent that stack every time.

Motivation: I wanted the homepage and about page to stay aligned with what I actually write and build — without manually rewriting hero copy, current focus, or philosophy every time I ship new content.

Problem: Home and about drift. You add writing entries and project entries; the main pages still say last quarter's focus or a generic tagline. Manually syncing is tedious and easy to forget. The corpus is the source of truth, but nothing was pulling from it into the locale JSON that drives the site.

Motivation: I wanted creating a .env for a new developer to be as easy as possible — no copy-paste guesswork, no wrong values that only surface in production, no tribal knowledge about which vars are required.

Problem: Onboarding usually means 'copy .env.example to .env and fill in the blanks.' That fails in predictable ways: people leave placeholders in, mistype URLs, forget which vars are required, and only discover mistakes when the app crashes or a feature silently breaks. There was no single CLI that walked you through setup with validation and guardrails built in.

Motivation: I wanted my AI agent to stop writing like a press release and start writing like me — same voice, same frontmatter, same structural patterns — without re-explaining those constraints every session.

Problem: Every time I asked an agent to write content for my site, the result came back hollow. Correct grammar, zero personality. The model defaulted to corporate filler because it had no constraints to work against. Re-prompting my voice preferences each session was unsustainable.

Motivation: I wanted to build a Super Mario–style platformer and stress-test a full spec-first, agent-driven workflow — spec-kit for requirements, Cursor vibe coding, and custom commands for self-review, testing, and memory — to see how far I could push it before having to step in.

Problem: Agent-built projects often ship with hidden bugs and no clear next step. I wanted to see whether a structured requirement spec plus agent commands for self-review, unit tests, e2e blackbox testing, and a SQLite-backed step memory could reduce that friction and eventually replace the need for a human operator.