kylebrodeur

Bootstrap a new or existing WordPress project with the full wp-agent-kit system: monorepo detection, Composer/WPackagist/SatisPress setup, Playground local dev, WP Engine dev/staging/production, CI gate, git hooks, and plugin packaging. Use when setting up a new WordPress project from scratch, onboarding an existing codebase, or wiring up a monorepo with multiple plugins/themes.

Use when integrating Gravity Forms with OpenAI via Gravity Connect: configuring OpenAI Fields (real-time pre-submission AI) or OpenAI Feeds (async post-submission processing), setting token ceilings, routing through OpenRouter, using GPT Image generation, voice-to-text transcription, or building AI-powered form workflows.

Use when working with Gravity Forms: installing or updating core and add-ons via wp gf CLI, managing forms and entries programmatically, exporting/importing form JSON for version control, testing notifications, configuring Cloudflare Turnstile anti-spam, or turning forms into MCP-compatible schemas for AI consumption. Also use for the GFAPI PHP class and hook/filter development.

Use when working with Gravity SMTP: installing, configuring, or troubleshooting WordPress transactional email delivery via SendGrid, Mailgun, Postmark, Brevo, SES, or custom SMTP. Critical: always verify version ≥ 2.1.5 to protect against CVE-2026-4020 which leaks API keys via unauthenticated REST endpoint.

Use for Gravity stack setup and orchestration: SatisPress/Composer configuration for the full plugin suite (Gravity Forms, SMTP, Connect, GravityView, Spellbook, Pods), reading the .cursorrules and llms.txt templates, or understanding how all products fit together. For product-specific work, load the individual skill instead.

Use when working with Gravity Wiz products: installing or managing Spellbook (the Gravity Wiz platform), Gravity Perks (48+ form enhancement plugins like GP Populate Anything, Nested Forms, Limit Submissions), or any individual perk. Also use for Gravity Shop (WooCommerce integration).

Use when working with GravityView or GravityKit products: creating Views to display Gravity Forms entries on the front-end, configuring search/sort/filter, entry approval workflows, edit-entry forms, role-based access, merge tags, shortcodes, or the GravityActions/GravityBoard extensions.

Use when working with the Pods Framework: creating or managing custom post types (CPTs), custom taxonomies, custom database tables (ACTs — Advanced Content Types), or custom fields via Pods. Critical for AI agents: always declare whether a Pod uses wp_postmeta or ACT storage before writing any DB queries, WP_Query, or pods_field() calls.

Use when creating, updating, testing, or viewing components and stories in obsidian-arrow-sandbox. Covers the full authoring workflow — building Arrow components with Obsidian classes and oas-* utilities, creating *.stories.ts files in stories/views/ or stories/components/ (not src/), the complete defineStories API (variants, children, status, kind, readableWidth, decorator, per-variant notes), viewing the story viewer at /components, keeping code DRY through sub-component extraction, and knowing when to reach for utilities vs. Obsidian's own classes.

Use when updating or maintaining an EXISTING scaffolded obsidian-arrow project — refresh managed files with create-obsidian-arrow refresh (scripts, docs, CI, tools, src/main.ts, src/utilities.css, test — never src/components/), update agent skills with pnpm skills:update, fix skills scoping for nested repos, and re-pull Obsidian styling.

Use when writing @arrow-js/core (v1.0.6) html template literals — reactive vs static (wrap reactive reads in arrow functions), full-value attribute binding where returning false removes the attribute, .property and @event binding, keyed lists, and async component(fn, { fallback }) via boundary(); plus the footguns — no literal HTML comments and no partial attribute values (both throw Invalid HTML position at render), and @event handlers must type the param as Event not a narrowed subtype like MouseEvent (TS2345).

Use when designing the component structure for a new view area or refactoring an existing one. Surveys the codebase for existing components and shared patterns, proposes a src/views/ + src/components/ hierarchy with file trees, asks targeted questions to confirm decisions, checks for DRY violations and improper primitive use, and produces a locked hierarchy doc the migration agent can execute. Run this BEFORE writing any component code or migration plan.

Use when building reactive Obsidian plugin UI with @arrow-js/core beyond basic templates — icons (Obsidian uses Lucide; the data-icon sweep since setIcon can't run inside templates), CSS scoping/specificity against Obsidian's global rules, component mount/unmount lifecycle in an ItemView, organizing shared reactive state, and imperative-DOM/floating-UI work (popovers, menus, positioning, focus, measuring). Complements arrow-js-obsidian-templates (syntax) with integration patterns. For the full CSS decision hierarchy (Obsidian classes → oas-* utilities → custom CSS) and token/class reference, see obsidian-arrow-css.

Use when porting an Arrow component from the obsidian-arrow-sandbox into the real Obsidian plugin, or wiring a check that the plugin copy hasn't drifted from the sandbox source. Covers the content-addressed porting-parity tool (scripts/component-hash.mjs), the canonical-form rules, a port-parity manifest, and a husky/CI pre-commit hook. The sandbox is the source of truth; the plugin copy must match.

Use when prototyping or building Arrow.js UI for an Obsidian plugin in the obsidian-arrow-sandbox project — covers running the sandbox, pulling Obsidian's real app.css from the local install, the dev/verify workflow, CSS scoping, and porting a finished component into a plugin's ItemView with near-zero refactoring. For the full component + story authoring workflow, see the obsidian-arrow-stories skill.

Use when styling Arrow.js components for an Obsidian plugin — deciding between Obsidian's semantic classes, oas-* utility classes, and custom CSS; reading the live token and class references; writing scoped rules that win the specificity battle without leaking; using CSS custom properties for overrides; and auditing components to minimize hand-written CSS.

Find copy-pasted code blocks and similar patterns using jscpd. Use when asked to find duplicate code, identify refactoring targets, or check if a utility should be extracted because it appears in multiple places.

Orchestrate a full codebase analysis using all static analysis tools (knip, dependency-cruiser, jscpd) and synthesize findings.

Find unused files, exports, types, and package dependencies using knip. Use when asked to find dead code, unused exports, or files that can be deleted.

Find copy-pasted code blocks and similar patterns using jscpd. Use when asked to find duplicate code, identify refactoring targets, or check if a utility should be extracted because it appears in multiple places.

Orchestrate a full codebase analysis using all static analysis tools (knip, dependency-cruiser, jscpd) and synthesize findings.

Find unused files, exports, types, and package dependencies using knip. Use when asked to find dead code, unused exports, or files that can be deleted.

1Password CLI integration for managing secrets, API keys, and authentication. Use when working with 1Password in Linux/WSL/Windows environments. Supports Service Account authentication, Desktop App integration, and Session Token Caching.

SSH agent management with 1Password for passphrase handling and keychain integration. Use when working with SSH keys on Linux/WSL/Windows. Features keychain-based key persistence and 1Password-stored passphrases for secure, passwordless SSH operations.

1Password environment management for AI agents and development. Load secrets as environment variables from 1Password, supporting .env and .env.1pass formats. Use when configuring API keys, tokens, or secrets for AI models, CI/CD, or local development.

Sync React components to Figma designs. Use when: 'sync component to Figma', 'generate Figma from code', 'create design system in Figma', 'export React to Figma'. Parses React/TSX with Babel AST, resolves Tailwind classes, outputs Figma-compatible JSON for loading via Figma Desktop plugin.

Visual-first UI/UX collaboration loop using agent-browser (primary), Playwright MCP (alternative), Figma MCP (design system/specs), and Lucid (wireframes). Use when designing or iterating on UI, reviewing the live app visually, creating wireframes, making layout decisions, discussing design before building, or running a design→build→verify loop. Trigger phrases: 'let's work on the UI', 'show me what it looks like', 'create a wireframe', 'design the layout', 'take a screenshot', 'browser view', 'before we build let's decide'.

Install and configure all ux-collab dependencies. Use when: setting up ux-collab for the first time, something isn't working, check.sh reports failures, agent-browser is missing, Playwright MCP is missing, ImageMagick is missing, or the user asks to 'set up ux-collab', 'fix the dependencies', or 'get ux-collab ready'.

Install, configure, verify, or remove the wsl-optimize toolset on a WSL2 machine, and harden the host against silent OOM shutdowns. Use when the user says "install wsl-optimize", "set up WSL memory protection", "my WSL keeps dying", "set up disk automation for WSL", "harden WSL", is provisioning a fresh WSL distro, or wants to uninstall. Covers the tools, the systemd timers, .wslconfig tuning, earlyoom, and persistent journald.

Diagnose and fix memory and disk exhaustion on a WSL2 dev machine that runs many AI coding agents. Use when WSL feels slow, runs out of memory or disk, when VS Code disconnects from WSL or the distro shuts down with no error, when the user says "WSL crashed", "WSL keeps dying", "free up space", "what's using my memory", "reclaim disk", "compact the vhdx", or for routine maintenance. Drives the wslreport, wsl-reclaim, wsl-compact, capmem, and wsl-optimize-doctor command-line tools with a safety-first workflow.

Install and configure the mac-optimize disk and shell hygiene toolset on a macOS machine, including the launchd automation (a low-disk watcher plus a weekly reclaim) and optionally moving a 1Password service-account token out of a cleartext dotfile into the login Keychain. Use when setting up a new Mac, when the user says "install mac-optimize", "set up disk automation", or "set up the disk cleanup tools", or when uninstalling them.

Diagnose and reclaim disk space and audit stray git worktrees on a macOS dev machine that runs many AI coding agents. Use when the Mac is low on disk, filling up, or feels slow; when the user says "clean up my mac", "free up space", "reclaim disk", "what's using my disk", "prune caches", or "find/clean git worktrees"; or for routine maintenance. Drives the diskreport, mac-reclaim, and worktree-audit command-line tools with a safety-first workflow.

Use when you want to hand a coding or research task to a second, persistent agent instead of doing it yourself — offloading long or independent work, running a cheaper/different model alongside your own, or keeping a separate context that accumulates across a series of related tasks.

Use when you need to run, operate, or troubleshoot the omp-rpc daemon — starting a persistent omp (oh-my-pi) RPC session, sending it tasks, switching its model live, steering or aborting a running turn, picking an Ollama Cloud model, reading its logs, or stopping and cleaning it up.

GitHub CLI extension for browsing starred repositories with AI-accessible JSON output

Universal append-only JSONL ledger with qmd semantic search and tiered HITL (strict, gated, autopilot). Use when tracking structured facts, decisions, requirements, research, or any append-only record across projects. Supports dynamic context injection into prompts. Install the pi-qmd-ledger extension first.

Create zero-duplication TypeScript field mappings with automatic type inference. Use when transforming between different object schemas (API responses to domain models, database rows to DTOs, external data to internal structures). Eliminates manual type definitions and ensures bidirectional type safety.

Bridges the `cap` command from a WSL/Linux session to a Windows-hosted Cap Desktop installation, since Cap screen capture must target the Windows desktop. Use when `cap` is not on PATH in WSL. Not needed on native macOS/Linux — Cap's own installer already places a native `cap` binary there.