A short tour of five macOS menu bar apps I built to scratch my own itch, and why AI agents, XcodeBuildMCP, XC testing, and remote Mac runners make this kind of tiny native tooling much easier to ship.
A practical index of Claude Code skill workflows: questioning, logo design, UI review, SEO research, and repo guardrails.
AskUserQuestion is the Claude Code tool that turns vague requests into structured decisions. Use it inside skills to force clarification before an agent writes code.
Dogfooding notes from a week of Mac menu bar apps, iOS workflows, BugDrop, issuectl, and small agent-powered developer tools.
A step-by-step guide to using Claude or ChatGPT for AI tax filing. Collect your documents, redact sensitive info, clean up the noise, and let AI help you fill out every form — without uploading your SSN to the cloud.
Stop re-authenticating every Playwright MCP session. This Claude Code plugin saves storageState authentication profiles per user role — log in once, keep your session persistent across sessions.
It's 2026 and big checks are still being cut to pre-product AI companies. Here's my standard reply to the recruiters.
The brute-force framing misses the point. For well-defined closure tasks — plan refinement, prototype validation, implementation verification — the Ralph Loop doesn't stumble toward correctness. It converges.
Nine layers of guardrails that make AI coding agents safe, predictable, and actually useful — from CLAUDE.md to CI pipelines to specialized review agents.
The old question was 'can you code?' The new question is 'can you build and ship?' A framework for evaluating engineers who use Claude Code, Cursor, and other AI-assisted coding tools.
