Skills

Gated SDLC loop with human approvals: intent, spec, plan, build, ship, maintain.

Build and run the app to prove a change works end to end. For agents without a built-in verify, such as Codex; in Claude Code use the bundled /verify.

One skill for PostgreSQL, MySQL, SQLite, and MongoDB: detect the engine, connect credential-safe, read the schema, read before writing, never write to prod.

Explain SQL and batch logic to a non-expert: numbered source lines, the real rows they read, per-line arithmetic, a results table where only the input changes, and an executed counter-example.

Find deepening opportunities, show them as an HTML report, then grill through the one you pick.

Build a throwaway prototype to answer a design question.

Run a retrospective on a coding session.

Drive Jenkins from the terminal with the jk CLI: jobs, runs, logs, artifacts, credentials.

Find and fix interaction bugs with seeded personas, invariant oracles, reproducible failures, and replay. App profiles fit it to games, CRUD apps (double submit, drafts, Korean IME, two sessions), AI chat (streaming, stop, safe rendering), and media (recording, permissions, interruptions). Godot, native iOS, and web adapters.

Review SwiftUI code for modern API use, data flow, navigation, accessibility, and performance, with before/after fixes. Targets the latest iOS and Swift; respects a project's existing deployment target.

One quality bar and improvement loop for apps, sites, and games: four dimensions (experience, function, consumer trust, engineering), hard floors, an independent judge for release candidates, and a ledger where every fix leaves a guard. Every report opens with a quality % (target 90% or higher) and why each point was lost.

Stop. That last message did not land: re-pitch it.

Compact the conversation into a handoff document for another agent to pick up.

How to write skills, AGENTS.md, and CLAUDE.md that agents follow.

Lazy-senior-dev mode: the simplest, shortest solution that actually works.

Design, critique, audit, and polish frontend interfaces.

44 diagram and chart types (architecture, flowchart, sequence, state, ER, timeline, Sankey, and more) as self-contained HTML/SVG/PNG, with Mermaid, .drawio, and .excalidraw import.

Generate and edit images with ChatGPT's current image model through your ChatGPT plan and the Codex CLI. It never names a model, so a newer ChatGPT image model works without changes. Modified here to keep working on future Codex releases.

Code-driven motion-graphics videos: keynote-style product launch videos built from real captures of your running app, and vertical YouTube Shorts, Reels, and TikTok.

Measurable UI/UX rules for iPhone apps, mobile web, and mobile games: touch targets, safe areas, contrast, Korean line breaks, HUD and combat readability. Ships a 7-category scoring rubric, an independent-judge prompt, a pre-ship checklist, and per-project overrides. Pairs with impeccable.

Hand a bounded task to a peer CLI: Codex, Claude, Antigravity, Kiro, NotebookLM, ChatGPT Pro.