Find Skills
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
Red Team
Perform dynamic agentic security testing of a codebase and its running services. Complements deep-review (static analysis) with behavioral anomaly detection, adversarial testing, invariant verification, and emergent behavior discovery. Produces a structured report with attack success rates, invariant violations, behavioral anomalies, and attack chains. Run AFTER deep-review for maximum coverage.…
Liquid Glass
Implement, adopt, review, or design Liquid Glass in SwiftUI, UIKit, or WidgetKit (iOS 26+). Use for glassEffect, GlassEffectContainer, glass buttons/toolbars/tab bars, morphing transitions, "make it glassy", refactoring a feature to Liquid Glass, or any API question about Liquid Glass materials, variants, and behaviors.
Ship
Plan and execute in one command. Runs /preflight, then /plan, then shows a compact plan digest for GO/ABORT confirmation, then runs /apply. The confirmation gate lets you catch a wrong-direction plan before committing to a full wave run. Pass --auto to skip confirmation. Pass --tidy to run /tidy --apply after execution. Pass --research to enable parallel research during planning. Pass --sequentia…
Checkpoint
Save or resume session state via a per-process .claude-checkpoint.<session>.md file. Reads the current checkpoint and reports status, or pass "save" as argument to write a new checkpoint. Per-process naming keeps concurrent Claude Code sessions on the same directory from clobbering each other's checkpoint. Use between sessions to preserve continuity.
Ux Review
Audit an EXISTING app's experience — not its pixels — against the CANON.md experience laws. Use when asked to "review the UX", "audit the flows", "why does this app feel annoying/clunky", "is it intuitive", count taps, check navigation architecture, find dead ends, or vet habit mechanics for dark patterns. Walks real journeys in the simulator (iOS) or the browser (web) counting taps or clicks and…
Skin Gallery
Debug-only component gallery screen for skin/theme iteration — every kit component under each candidate skin side by side, launch-seam reachable, light+dark screenshot evidence, contrast-audited, versioned by iteration. Use when choosing or iterating app colors/themes/skins, or when a reviewer needs to judge palette candidates.
Apply
Execute docs/plan.md across any function (engineering, marketing, sales, operations, strategy). Defaults to --pool mode (multi-session coordination via atomic git-ref claims, ADR 013); pass "--static" to force single-session wave dispatch or "--sequential" to force single-task mode. When the `example-app` CLI is on PATH, ENGINEERING tasks are EXECUTED by example-app (ADR 015, apply/example-app-EX…
Ux Design
Design the user EXPERIENCE before any screen is drawn — user journeys, flow architecture, navigation model, state machines, and ethical habit loops, for iOS or web. Use when asked to "design the UX", "design the flows/journeys", "make it intuitive", "make it habit-forming", "how do users get things done", map navigation, or when screens exist but feel disconnected (taps go nowhere). Produces jour…
Preflight
Pre-wave safety check. Validates git state, disk space, stale worktrees, build health, and plan readiness before spawning parallel agents. Outputs a GO/NO-GO summary. Run before any /apply wave to prevent the class of failures that require mid-wave manual recovery.
Ui Review
Review the UI of an EXISTING app (iOS or web) from its real rendered pixels plus its source. The looks half of the pre-ship audit; /ux-review is the works half. Use when asked to "review the design", "design audit", "how does this app look", "critique the UI", "why does this feel off", or to grade an app against its design system. Captures light+dark screenshots per core screen (simulator on iOS,…
Security Audit
Complete security audit combining static analysis (deep-review) and dynamic testing (red-team). Runs deep-review first, then red-team, producing two complementary reports. Use with an optional scope as argument.
Web Craft
Production web craft — implement or harden a web UI so it feels engineered rather than generated. Covers token discipline, responsive verification across viewports, keyboard and focus behavior, URL-as-state, layout-shift and performance budgets, form and error craft, and reduced-motion. Use when implementing an approved web design, reviewing or polishing React/Next/Vue/HTML UI code, fixing UI tha…
Eventkit
Create, read, and manage calendar events and reminders using EventKit and EventKitUI. Use when adding events to the user's calendar, creating reminders, setting recurrence rules, requesting calendar or reminders access, presenting event editors, choosing calendars, handling alarms, observing calendar changes, or working with EKEventStore, EKEvent, EKReminder, EKCalendar, EKRecurrenceRule, EKEvent…
Go
Conventions for writing and reviewing Go code — error handling, testing, package structure, concurrency, and the zero-stub policy. Use when implementing, refactoring, or reviewing Go in any repo; load before writing the first line, not after.
Dogfood
Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to "dogfood", "QA", "exploratory test", "find issues", "bug hunt", "test this app/site/platform", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue --…
App Store
App Store work in three modes. OPTIMIZE (ASO): keywords, keyword research, app name/subtitle, product pages, screenshots, in-app events, ratings, review prompts, localization. PREPARE: app store submission review-prep, app rejection reasons, App Store guidelines, PrivacyInfo.xcprivacy privacy manifest, required-reason APIs, ATT, IAP/StoreKit review rules, entitlements, metadata compliance, HIG ch…
Tidy
Unified cleanup skill. Harvests stranded work from branches/worktrees, digests scratch files into the tier system, trims completed work from the plan, and sweeps ephemeral artifacts. Dry-run by default. Use with --apply to execute. Scope with --harvest, --digest, --trim, or --sweep to run a single phase. Pass --prune to delete fully-merged branches. Pass --aggressive to also sweep dated reports.
Design
Design front door — interview first, then route to the right design skill. Use when the user says "/design", "design something", "I want to design an app/screen/feature", "make this beautiful", "which design skill should I use", or gives a design request too vague to route (no clear target, stage, or deliverable). Asks 3-4 multiple-choice questions (what/where/constraints/depth), then invokes the…
Plan
Create or update a granular, prescriptive work plan in docs/plan.md with checkable task breakdowns, dependency tracking, ADR generation, and use case discovery. Every task links to the user-facing use cases it enables. Use with a project description as argument, or a custom path and description (e.g., "path/to/plan.md description"). Supports --team or --research for parallel research before plann…
Coverage
Guarantee programmatic test coverage of every capability in the product. Resolves the capability/use-case catalog, builds a coverage matrix mapping each capability to its API and browser tests by running them (never self-reported), generates the missing tests in parallel (agent-browser authors, Playwright/API tests certify), and enforces a launch-readiness gate that is GREEN only when every capab…
Design Fidelity Loop
End-to-end method for getting real design quality out of agent-built UIs — DRY components first, project design rules as checkable tokens, in-framework gallery + screenshot iteration with the designer, skin/token propagation, founder-gated beautify waves. Use when an app's UI "works but looks agent-built" or when starting any design-quality program.
Ios Simulator
Manages iOS Simulator devices and tests app behavior using xcrun simctl. Covers device lifecycle (create, boot, shutdown, erase, delete), app install and launch, push notification simulation, location simulation, permission grants via privacy subcommand, deep link testing via openurl, status bar overrides, screenshot and video recording, log streaming with os_log filtering, get_app_container path…
Design Fidelity
Transform mockups/design canvases into production UI (SwiftUI or any framework) with automated design-fidelity validation — measured pixel predicates, two-sided self-tests, and a visual merge gate. Use when implementing approved designs, verifying an app matches its mockups, or authoring machine-checkable design-adherence predicates for agent-driven (e.g. example-app) UI work.
Handover
Wind a session down cleanly and hand its work to future sessions -- finish or park in-flight work, sweep your PRs, push WIP to a remote handover branch, write pickup notes, release claims, update the roadmap/board, and leave the repo in a state a fresh session can continue from. Use when ending a session with unfinished work, when told to "hand over", "wind down", "wrap up the session", before a…
Component Kit Consolidation
DRY a UI codebase into a shared component kit — grep-based duplication inventory, spec with dedup counts, invisible-refactor extraction waves with pixel-diff proof, rule of two, purge-list awareness. Use when asked to "DRY the components", consolidate duplicated UI, or before a design-quality program.
Agent Browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to…
Journal
Record investigation findings, debugging sessions, or benchmark results in docs/devlog.md. Use when you discover something worth remembering. Pass a description of the finding as the argument.
Resilient Agent Lanes
Run worker-agent lanes that survive unstable API streams and wedged sessions — tiny committed increments, on-disk heartbeats, resume-not-restart, respawn thresholds, background builds, resource leases, takeover protocol. Use when leading parallel agents, when agent sessions keep dying/stalling, or before dispatching a multi-agent wave.
Go Release
Set up GoReleaser and release-please GitHub Actions pipeline for a Go project with semantic versioning. Use with flags as argument: --lib for libraries, --bin for binaries, --org and --repo to specify the GitHub repository. Use when bootstrapping automated releases for a Go project.
Deep Review
Perform a multi-hour deep dive security, architectural, and functional review of a large codebase. Reads and traces actual code end-to-end, producing a structured report with file:line references, CWE/CVSS classifications, attack narratives, fix code, and prioritized remediation. Use with an optional scope as argument (e.g., "auth system", "api/", or no argument for full codebase).
Lint
Lint the knowledge WIKI (docs/lore.md, docs/devlog.md, docs/plan.md, docs/adr/, memory/) for contradictions, stale claims, orphan references, missing entries, and missing cross-references. This is the documentation/knowledge linter, NOT a code linter -- it never runs gofmt, eslint, ruff, or any source-code tool. Use when asked to audit, lint, sanity-check, or find drift in the project wiki or kno…
Verify
Audit a codebase or single feature end-to-end. Discovers architecture, catalogs use cases, checks wiring, tests that use cases work, then iteratively plans remediations, applies them, and re-verifies until the system is healthy or no further progress can be made (max 5 iterations). Pass --audit-only to stop after the report. Pass --auto to skip confirmation gates.
Humanizer
|
Concise
>
Widgetkit
Implement, review, or improve WidgetKit widgets and controls. Use when building Home Screen, Lock Screen, StandBy, or CarPlay widgets with timeline providers; configurable widgets with AppIntentTimelineProvider; interactive widgets or Control Center controls with Button/Toggle wiring; WidgetKit push reloads, refresh budgets, deep links, Smart Stack relevance, Liquid Glass/accented rendering, widg…
Honest Verdicts
Anti-vacuous-green playbook for test/CI evidence — named-suite + executed-count verdicts, bundle-membership checks, fresh-environment runs, platform-gating, Release builds, self-passing gates. Use when claiming tests pass, authoring CI gates, writing acceptance predicates, or diagnosing "flaky" CI.
Journey Qa
Journey-map-driven QA for agent-built UIs -- evidence-cited journey map (per-step WORKS/PARTIAL/STUB/MISSING), generated human test guide, PR journey-id rule, device-walk gate before distribution. Use when building/reviewing/shipping UI against user journeys, planning QA for an app, before any external distribution, or when asked "does the app actually work end to end".
Web Design
Design a web app or site in its own framework — direction anchors, design tokens, a component gallery, then full screens rendered headlessly at three viewports in light and dark for screenshot iteration. Use when designing a new web UI, landing page, dashboard or console, exploring visual directions for the web, asked for "web mockups", redesigning a React/Next/Vue/HTML surface, or starting a web…
Swiftui Craft
Production SwiftUI craft and native iOS/HIG compliance — make an app feel native and match Apple Human Interface Guidelines: spacing grid, typography and Dynamic Type, semantic colors and dark mode, safe areas and Dynamic Island, tab bars/sheets/modals, haptics, iOS accessibility (VoiceOver), SF Symbols, WidgetKit visual rules, and iPhone/iPad layout adaptation. Use when reviewing or improving Sw…
Activitykit
Implement, review, or improve Live Activities and Dynamic Island experiences in iOS apps using ActivityKit. Use when building real-time updating widgets for the Lock Screen and Dynamic Island — delivery tracking, sports scores, ride-sharing status, workout timers, media playback, or any time-sensitive information that updates in real time. Also use when working with ActivityKit, ActivityAttribute…
Ios Design
Claude Design for iOS — run the direction → tokens → component cards → full-screen mockup workflow directly in SwiftUI, no HTML. Renders light+dark PNG mockups headlessly (ImageRenderer, no simulator, no Xcode project) for screenshot iteration with the user. Use when designing a new iOS app or feature, exploring visual directions, asked for "mockups in SwiftUI", or starting a design system before…