cuozg
97 verified skills197 total stars
goal-todo
Convert a spec document into an implementation TODO list in the same spec folder. U se when the user says goal-todo, todo from spec, generate tasks from spec, turn this spec into todos, create implementation checklist, extract tasks, or asks to read a Docs/Specs design doc and produce what must be implemented. Includes UI/UX review and codebase investigation before writing the checklist. Do not use for implementing the tasks, creating new goal files, writing test cases, or verifying completed work.
development4
unity-standards
Shared Unity reference hub for writing, reviewing, debugging, planning, testing, documenting, and optimizing Unity C# systems.
development4
unity-technical
Create Unity technical solution documents from user requirements, feature ideas, bug goals, specs, or codebase problems. Use when the user asks for a technical approach, architecture, implementation strategy, solution options, feasibility analysis, system design, or "how should we build/fix this" for Unity runtime, Editor, tools, assets, data, UI, WebGL, SDKs, or production pipelines.
tools4
git-pr
Use this skill to generate and apply a clear title and structured description for a GitHub Pull Request by deep-investigating all changed files. Use it when the user opens a new PR, when a PR's title or description is missing or stale, or when the user says "update this PR", "describe this PR", or "write a summary for this PR." Applies the result directly via gh pr edit.
documentation4
unity-mcp-orchestrator
Orchestrate Unity Editor via MCP (Model Context Protocol) tools and resources. Use when working with Unity projects through MCP for Unity - creating/modifying GameObjects, editing scripts, managing scenes, running tests, or any Unity Editor automation. Provides best practices, tool schemas, and workflow patterns for effective Unity-MCP integration.
tools4
goal-execute
Execute one goal file created by goal-create. Use when a user says to run, implement, finish, or ship a goal in Docs/Goals, or wants one goal taken from pending to verified completion. Do not use for creating goals or read-only audits.
development4
unity-review-pr
Use this skill to review Unity GitHub pull requests. Use when the user provides a PR URL/number, or asks to "review PR", "pull request", "merge", "is this ready".
tools4
unity-investigate
Investigate Unity codebases — from quick Q&A to comprehensive system analysis reports. Auto-triages request complexity: simple questions (how does X work, what calls Y, trace this flow) get direct inline answers; complex requests (system analysis, pre-refactor audit, architecture investigation, team documentation) produce a structured markdown report with Mermaid diagrams, cited evidence, and risk tables. Use whenever the user asks "how does X work," "what calls this method," "trace the flow," "explain this system," "investigate this architecture," "do a deep investigation," "I need an investigation report," "full analysis of this system," or any question about codebase structure and relationships. Also use when the user wants to understand an unfamiliar subsystem before modifying it.
development4
goal-create
An interactive goal creation, update, and tracking skill that writes structured goal files to Docs/Goals/{feature-name}/{kebab-case-task}.md with acceptance criteria. Use for defining NEW goals, updating EXISTING goals, or scoping vague requests into concrete goals.
development4
unity-image-gen
Generate Unity raster image assets through Unity MCP: game sprites, item art, backgrounds, UI icons, portraits, concept images, transparent cutouts, image edits, upscales, background removal, and Unity scene or Game View screenshots. Use when a Unity project needs image files imported under Assets or screenshots captured from the editor. Do not use for meshes, audio, animation, materials, gameplay code, UI Toolkit layout, or generic non-Unity image generation.
tools4
goal-verify
Use when verifying Docs/Goals files, validating goal-execute results, checking acceptance criteria, or deciding whether a goal can be marked completed.
development4
unity-debug
Diagnose Unity bugs from compile errors to stack traces, runtime failures, and tricky logic issues; gather reproduction details, trace the code path, prove the root cause with evidence, and report options without changing code unless asked.
development4
unity-document
Unity documentation skill for investigating a feature in a Unity codebase and writing an evidence-based system document to Docs/Systems/<feature-name>.md.
development4
unity-debug-log
Generate formatted Debug.Log snippets for tracing values, events, or methods in Unity C# without modifying project files directly.
development4
unity-costing
Create tshirt costing, detail costing, highlevel costing, feature costing, task breakdown, estimate, effort sizing, or a detailed Unity delivery plan before implementation. Use for small-to-XL Unity work when a costed task tree is the deliverable; investigate the codebase first and fire subagents when scope touches multiple systems or evidence is needed.
development4
unity-sdk-integration
Maintain Unity SDK packages, package boundaries, editor integration, runtime initialization, service management, samples, integration tests, and third-party SDK boundaries. MUST use for `com.zeno.sdk`, SDK editor menus, package API compatibility, samples, initialization validation, and package-level integration checks. Preserve public SDK APIs unless breaking changes are explicitly requested.
development2
unity-addressables-content
Manage Unity Addressables content strategy, groups, labels, asset references, remote catalogs, content builds, and content update validation. MUST use for Addressables setup, group or label organization, remote content delivery, catalog/update checks, runtime load-risk review, or deciding whether assets should become Addressables. Do not use for installing packages without delegating to unity-package-manager, or for unrelated runtime loading code.
development2
goal-improve
Improve goals by critically reviewing whether they are useful, well designed, pragmatic, architecturally sound, compelling, intuitive for humans, and executable by AI coding agents. Use before goal-execute when goals feel vague, overbuilt, low-value, poorly sequenced, hard to verify, or when the user asks what you actually think of the goals.
development2
unity-app-ui-mvvm
Expert for App UI MVVM pattern and dependency injection - ObservableObject, RelayCommand, AppBuilder, service registration, and DI. Use this skill whenever the user wants to create a ViewModel, bind data to UI elements, set up dependency injection, register services, implement commands for button actions, use [ObservableProperty] or [RelayCommand] attributes, create an AppBuilder, or structure their Unity App UI project with the MVVM architecture pattern. Also trigger when the user asks about data binding in UXML, property change notifications, or service lifetimes (transient, singleton, scoped).
development2
unity-app-ui-redux
Expert for App UI Redux state management - Store, Slices, Reducers, Actions, AsyncThunks, middleware, and Redux DevTools. Use this skill whenever the user wants to manage global state, create a Redux store, define actions and reducers, dispatch events, subscribe to state changes, implement async data fetching with thunks, add middleware for logging or analytics, use the Redux DevTools window, or architect a predictable state management layer in their Unity App UI project. Also trigger when the user mentions ActionCreator, StoreFactory, PartitionedState, or asks about centralized state, immutable records, or state slices.
tools2
flutter-init
Initialize Flutter project structure from user requests: feature-first folders, pubspec.yaml with Riverpod/GoRouter/Dio/Freezed, analysis_options.yaml, .gitignore, and boilerplate Dart files under lib/. Use when starting a new Flutter project, setting up folder structure, creating a template, organizing a project from scratch, or when the user gives an org name, project name, and feature list. Triggers include folder layout, project organization, directory structure, project template, pubspec setup, best folder structure for Flutter, flutter create with architecture, and set up a Flutter project with Riverpod. Do not use for adding features to an existing project; use flutter-code for that.
development2
unity-app-ui
Expert assistant for developing UI in Unity using App UI framework. Use when creating components, styling, MVVM architecture, navigation, or any UI-related tasks. Trigger this skill whenever the user mentions App UI, Unity UI Toolkit components like Panel/Button/TextField, UXML layouts with appui namespace, USS styling with appui variables, overlays (popover, modal, toast, menu), or building any Unity user interface. Also trigger when the user asks to create a form, dialog, settings screen, dashboard, or any visual component in Unity.
tools2
unity-audio
Generate, edit, clean up, loop, normalize-by-factor, and validate Unity AudioClip assets with Unity MCP tools. MUST use for placeholder SFX, music loops, ambience loops, trimming silence, trimming ranges, volume changes, loop processing, and audio asset preparation. Do not use for gameplay audio systems or mixer code unless the work is primarily AudioClip preparation.
tools2
unity-init
Use this skill to initialize a Unity project structure based on user requests: scaffolding folder layout with feature-based organization, assembly definitions (.asmdef), .gitignore, and namespace-aligned directories under Assets/_Project/. Use whenever a user wants to start a new Unity project, set up folder structure, initialize or scaffold a project, create a project template, or organize their Unity project from scratch, even if they just say they are starting a new Unity game or describe a game concept and ask for the project setup. Also triggers on folder layout, project organization, directory structure, project template, asmdef setup, or best folder structure for Unity. Always use this skill when the user provides a company name, project name, and feature list for a new Unity project. Do not use for adding features to an existing project; use unity-code for that.
development2
unity-input
Work on Unity input across gameplay, UI, controllers, rebinding, touch, and debugging. MUST use for Input System action assets, legacy Input Manager checks, PlayerInput components, generated input wrappers, controller support, mobile touch controls, rebinding UX, and input behavior validation. Do not convert input backends or replace bindings without explicit approval.
development2
unity-package-manager
Inspect, install, remove, embed, unembed, and import samples for Unity packages through Unity MCP Package Manager tools. MUST use for package metadata checks, manifest/lock compatibility, adding or removing registry packages, embedding local packages, unembedding packages, and sample import. Never change packages without explicit user approval.
tools2
unity-qa-automation
Automate Unity QA checks with MCP inspection, scene hierarchy review, GameObject/component validation, Play Mode smoke checks, screenshot or scene captures, console log scanning, and regression evidence. MUST use for non-destructive automated verification, smoke checks, scene validation, screenshot validation, and pass/fail reports. Do not replace unit tests, manual QA test cases, debugging, or profiler analysis.
tools2
unity-project-settings
Safely inspect and modify Unity project-level settings including tags, layers, build settings, Player Settings, quality settings, package/project configuration, and editor settings. MUST use for high-impact project configuration changes that require read-before-write discipline and console verification. Avoid bulk settings rewrites when targeted changes are possible.
development2
unity-scene-builder
Build, modify, inspect, and validate Unity scenes using Unity MCP scene, GameObject, asset, and capture tools. MUST use for scene creation, loading, hierarchy setup, object placement, component wiring in scenes, camera/light/layout validation, and visual scene captures. Avoid destructive scene edits and save only when explicitly requested.
tools2
unity-mcp-automation
Standardize safe, repeatable use of Unity MCP tools across scripts, scenes, GameObjects, assets, packages, editor state, console checks, commands, captures, and validation. MUST use as a meta-skill for MCP-heavy Unity tasks spanning multiple tool categories. Do not replace specific skills such as unity-code, unity-debug, unity-scene-builder, unity-asset-generation, or unity-package-manager when one clearly owns the task.
tools2
unity-mobile
Validate and troubleshoot Unity Android and iOS projects, dependency resolver issues, Gradle settings, permissions, push notifications, ads, mobile SDK integration checks, and mobile build readiness. MUST use for mobile-specific Unity problems involving `Assets/Plugins/Android`, Mobile Dependency Resolver, Push Notifications, LevelPlay, platform permissions, or device-only behavior. Do not modify manifests, Gradle files, signing, or resolver config without approval.
tools2
unity-save-data
Design, inspect, implement, and validate Unity save/load systems, serialization formats, migrations, versioning, local/cloud-save boundaries, corrupt-data handling, backup safety, security, privacy, and backward compatibility. MUST use for persistent player data and SDK-managed state. Never delete, migrate, or overwrite saved data without explicit approval.
development2
goal-loop
Loop all goals. Scan the goal registry for incomplete goals; when one is found, improve weak goal definitions with goal-improve when needed, execute it with goal-execute, verify it with goal-verify, then loop again until no incomplete goals remain. Picks one goal per cycle deterministically by priority and never exits on an unverified completion claim. Use for unattended batch workflows.
development2
unity-liveops
Plan, inspect, and validate Unity LiveOps integrations for remote config, analytics, ads, IAP, economy tuning, events, A/B testing, push notifications, consent, and release-safe operations. MUST use for live game service readiness, SDK service initialization order, test-mode checks, privacy gates, and production configuration reviews. Do not add secrets or production keys to source-controlled files.
development2
plan-test
Verify if goals have been implemented by checking acceptance criteria against codebase. Searches repo for code/test evidence, generates structured test reports with verdicts. Use after plan-work to validate implementation. Auto-triages quick vs deep mode.
development2
unity-plan
Use this skill for Unity cost estimation. Make sure to use it whenever the user asks what a Unity feature, refactor, migration, or integration will cost, how long it will take, how much work it is, wants an estimate, effort breakdown, risky parts, or phased delivery for pricing — even if they never say "costing". Investigate the codebase first with parallel `explore` subagents, use a `librarian` subagent only when external packages, SDKs, or services affect the estimate, then return evidence-backed hour ranges, task breakdowns, assumptions, risks, and confidence. Do not use it for implementation, design docs, or generic Unity planning without estimate intent.
development2
html-prototype
Create interactive HTML app prototypes and landing pages as a single self-contained HTML file with full-screen layouts, multi-screen navigation, smooth transitions, clickable controls, fake auth/data, and product-aware visual design. MUST use when the user describes an app idea and wants to see it, prototype it, mock up screens, build a clickable demo, landing page, marketing site, product page, interactive wireframe, app demo, or any multi-screen flow with navigation. Also use for app concepts like recipe, fitness, dashboard, SaaS, e-commerce, social, chat, finance, booking, wellness, or education when the user wants to visualize the flow or test screen transitions. Do not use for static diagrams, production code, or game prototypes.
tools2
imagegen
Use this skill when the user asks to generate, create, or edit images — concept art, product shots, icons, logos, textures, illustrations, or batch variants. Covers generation from text prompts and editing existing images (background removal/replacement, inpainting, style transfer, object removal). MUST use when the user says "generate an image," "create an icon," "make me a picture," "I need a hero image," "swap/remove the background," "edit this photo," "make it look like [style]," "batch of icons," "placeholder art," or describes needing any visual asset casually. Also triggers on: game assets (sprites, textures, UI icons), website visuals (hero images, blog headers), product photo edits, logo generation, infographics, or any request to produce an image from a description. Runs the bundled CLI (scripts/image_gen.py) and requires GEMINI_API_KEY or GOOGLE_API_KEY.
tools2
unity-spec
Investigate Unity projects and generate per-feature Game Design Specification docs in Docs/Specs/. Supports Full GDD, single Feature Spec, and Update mode that diffs existing specs against the codebase. MUST use when the user asks for a game spec, GDD, game design document, feature specification, stale spec update, per-feature spec split, or structured design documentation for a game concept. Also use to document mechanics, systems design, progression, UX/UI flows, or technical constraints. Do not use for technical design docs or system docs (unity-document), task planning (unity-plan), codebase investigation without spec output, or project scaffolding.
development2
plan-loop
Autonomous orchestrator for batch goal completion. Loops through all incomplete goals, executing plan-work → plan-test → plan-improve cycles until all goals are verified complete. Picks one goal per cycle deterministically by priority. Detects gaps, invokes fixes, re-tests, and repeats. Generates completion report. Use for unattended batch workflows ("run all goals", "autonomous completion", "loop until done").
development2
unity-app-ui-theming
Expert for App UI theming and styling - USS variables, custom themes, dark/light mode, scale factors, and BEM conventions. Use this skill whenever the user wants to change colors, create a branded theme, customize fonts or typography, adjust spacing, implement dark/light mode switching, toggle between themes at runtime, set up responsive scaling (small/medium/large), register custom icons, style components with USS variables, apply BEM naming conventions, create a .tss or .uss file, or configure PanelSettings for theming. Also trigger when the user mentions design tokens, appui-- prefix, ThemeContext, ScaleContext, or asks about color palettes and USS variable overrides.
testing2
unity-build-pipeline
Prepare and validate Unity builds across Android, iOS, desktop, and CI. MUST use for build readiness checks, Player Settings validation, scenes-in-build review, scripting defines, package resolution, platform requirements, CI build configuration, and non-WebGL build pipeline issues. Delegate WebGL-only work to unity-webgl and do not change platform/signing settings without explicit approval.
development2
flutter-standards
Use this skill as the shared reference hub for all Flutter & Dart development — coding standards, naming conventions, architecture patterns, state management, testing, performance, and UI guidelines. MUST be included in load_skills for any Flutter task delegation. Triggers automatically when writing, reviewing, debugging, testing, or planning Flutter/Dart code. Contains 12 reference files across architecture, state, UI, performance, and tooling categories that downstream skills pull on demand. Also use when the user says Dart style guide, Flutter best practices, Riverpod patterns, feature-first structure, Flutter performance, or any Flutter/Dart coding standards question.
tools2
unity-localization
Implement, inspect, and validate Unity localization for locale setup, string tables, localized assets, font fallback, RTL handling, UI overflow, missing translations, and localization QA. MUST use for Unity Localization package work, multilingual UI/content checks, fallback behavior, and text coverage validation. Delegate package setup to unity-package-manager when installation or removal is required.
testing2
unity-prefab
Create, modify, duplicate, wire, inspect, and validate Unity prefabs through Unity MCP tools. MUST use for prefab creation, component setup, serialized field references, prefab variants or duplicates, asset path validation, and non-destructive prefab modification. Prefer explicit asset paths over ambiguous object names and preserve existing prefabs unless modification is requested.
tools2
html-prototype
Create interactive HTML app prototypes and landing pages as a single self-contained HTML file with full-screen layouts, multi-screen navigation, smooth transitions, clickable controls, fake auth/data, and product-aware visual design. MUST use when the user describes an app idea and wants to see it, prototype it, mock up screens, build a clickable demo, landing page, marketing site, product page, interactive wireframe, app demo, or any multi-screen flow with navigation. Also use for app concepts like recipe, fitness, dashboard, SaaS, e-commerce, social, chat, finance, booking, wellness, or education when the user wants to visualize the flow or test screen transitions. Do not use for static diagrams, production code, or game prototypes.
tools2
unity-animation
Create, convert, clean up, trim, loop, and wire Unity animation assets through Unity MCP tools. MUST use for sprite-sheet animation clips, humanoid animation cleanup, stationary/root-motion conversion, loop trimming, generated humanoid animation, AnimatorController creation, and validation of `.anim` or `.controller` assets. Do not use for general runtime animation code unless the task is primarily asset preparation.
tools2
unity-app-ui-navigation
Expert for App UI navigation system - NavGraph, NavHost, NavController, destinations, and visual controllers (AppBar, Drawer, BottomNavBar). Use this skill whenever the user wants to add screens or pages to their Unity app, create a navigation graph, set up a bottom tab bar, implement a side drawer or hamburger menu, add an app bar or toolbar, navigate between views, pass data between screens, manage back stack behavior, create a multi-screen flow (onboarding, login, settings), or implement any kind of screen-to-screen navigation in App UI. Also trigger when the user mentions NavigationScreen, NavigationRail, NavAction, or asks about deep linking or nested navigation.
tools2
plan-improve
Quality refinement engine — reads Docs/Goals/**/*.md (recursively, including feature subfolders), assesses work output against acceptance criteria, identifies gaps, delegates targeted improvements, and verifies results. Use after plan-work completes, when the user says 'improve this,' 'make it better,' 'check against goals,' 'refine the work,' 'quality pass,' 'plan improve,' or wants post-execution quality review. Runs autonomously like plan-work but focused on QUALITY over COMPLETION.
development2
plan-goal
Interactive goal creation, update, and tracking skill — write structured goal files to Docs/Goals/{feature-name}/{kebab-case-task}.md with acceptance criteria, and maintain Docs/Goals/Master.md as the central goals registry. Use for defining NEW goals, updating EXISTING goals (change priority, add criteria, mark complete/blocked), viewing goals dashboard, or scoping vague requests into concrete goals. Triggers: 'new goal,' 'create a goal,' 'plan this,' 'I want to build X,' 'break this down into goals,' 'scope this feature,' 'change the priority,' 'add a criterion,' 'mark as completed,' 'bump priority,' 'show me all goals,' 'goals dashboard,' 'what's the status.' Also use when the user describes a feature wish or improvement need that requires goal definition before implementation (e.g. 'make X faster,' 'we need Y,' 'add Z to the app'). Do NOT use for executing goals (plan-work) or reviewing completed work (plan-improve).
development2
unity-compile-check
Run Unity batchmode compile checks and parse Unity Editor log files for errors. MUST use whenever an agent needs to verify C# code compiles in Unity — this is the ground-truth compilation check. Use after writing scripts, refactoring, fixing bugs, changing asmdef references, updating packages, or before committing multi-file changes. Use to parse existing Unity log files from CI or build machines. Triggers: "does it compile," "compile check," "verify compilation," "Unity log," "parse log," "batchmode," "compile errors," "check for errors," "headless compile," "run Unity compiler," "CI compile failed." ALWAYS use when lsp_diagnostics is not enough — cross-assembly errors, platform defines, asmdef boundaries, domain reload issues. Do not use for runtime debugging (unity-debug), performance (unity-optimize), or writing code (unity-code).
development2
cloud-infra
Unified cloud infrastructure skill for Cloudflare R2 storage, Vercel deployment, signed URLs, CDN caching, and file upload patterns for Next.js apps. Auto-triages Quick work like one config, upload endpoint, or cache header fix, and Deep work like full upload pipelines, CDN strategy, or R2+Workers integration. MUST use for cloud infrastructure requests: R2 bucket setup, presigned URLs, Vercel config, CDN caching, file uploads, image optimization, edge caching, cache invalidation, storage integration, Cloudflare Workers, S3-compatible storage, Cache-Control, ISR revalidation, stale-while-revalidate, wrangler, or R2 bindings. Do not use for database design, API route logic, or frontend components.
development2
plan-work
Pick one incomplete goal, implement it end-to-end, and verify no errors after implementation. Autonomous single-goal execution pipeline for Docs/Goals goal files. Runs goal through fixed specialist phases: discover goal, plan, review, implement in isolated git worktree with PR, independently verify every acceptance criterion with concrete evidence (file:line, test results, command output), then sync goal docs, Master.md, README, and specs. Use when the user says 'execute goal', 'run goal', 'plan work', 'do the goal', 'start working', 'execute the plan', 'implement everything', 'autonomous mode', 'just do everything', or `/omo/work`. Also use when Docs/Goals contains goals and the user wants unattended completion of one goal. Do NOT use for creating goals; use plan-goal. Do NOT use for quality refinement; use plan-improve. Do NOT use for verification-only reports; use plan-test.
development2
skill-creator
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
testing2
unity-webgl
Unified Unity WebGL skill for JavaScript interop plugins, WebGL build optimization, deployment, and custom HTML templates. Auto-triages JSLib, Build, and Template work. MUST use for Unity WebGL requests: .jslib plugins, calling JS from Unity, browser APIs, build size, compression, memory, linker.xml, Player Settings, deployment, server MIME types, loading screens, templates, responsive canvas, iframe embedding, localStorage, Brotli, platform-conditional code, or issues that work in Editor but fail in WebGL. Do not use for general runtime C#, editor scripts, or non-WebGL debugging.
tools2
pixijs-coder
Write, review, debug, or explain PixiJS v8 code: applications, scenes, rendering, assets, shaders, filters, particles, animations, and performance. Auto-triages Quick tasks like one object, feature, or API question, and Deep tasks like multi-system apps, custom renderers, shader pipelines, or performance audits. MUST use for any PixiJS request: app setup, textures, Graphics, sprites, filters, shaders, draw-call optimization, v7/v6 migration, WebGL/WebGPU issues, PixiJS, PIXI.Application, Container, Ticker, Assets.load, particle systems, RenderGroup, or help with a pixi app. Always use v8 APIs unless explicitly asked otherwise.
development2
mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
tools2
unity-code
Unified Unity runtime C# coding skill — write, extend, or refactor runtime code using code-standards as the single source of truth for patterns and conventions. Integrates with Unity MCP for mandatory console verification after every code change. Auto-triages: Quick (single-file MonoBehaviour, ScriptableObject, interface, enum, struct, helper), Deep (multi-file features, services, state machines, refactors spanning 2+ classes). MUST use for ANY Unity runtime code request — creating scripts, adding components, implementing features, building systems, refactoring, or cleanup. Triggers: "write a script," "add a component," "build a system," "implement feature," "refactor," "clean up," "simplify." Do not use for performance optimization (unity-optimize), Editor scripts (unity-editor), UI Toolkit (unity-uitoolkit), tests (unity-test-unit), or debugging (unity-debug).
tools2
unity-asset-generation
Generate, convert, and prepare Unity assets through Unity MCP asset-generation tools. MUST use for Unity asset creation requests: sprites, sprite sheets, images, icons, placeholder art, materials, terrain layers, meshes, sounds, humanoid animations, animation clips, animator controllers, model texturing, retopology, rigging, or converting existing textures into Unity-ready assets. Distinguishes generative tools from non-generative conversion tools and validates generated assets in the Unity project. Do not use for writing runtime gameplay code, custom editor tools, UI Toolkit screens, or performance profiling.
tools2
phaser-coder
Write, review, debug, or explain Phaser 3 code: game config, scenes, sprites, Arcade or Matter physics, tilemaps, animations, audio, input, cameras, tweens, particles, groups, and performance. Auto-triages Quick tasks like one feature, API question, or bug fix, and Deep tasks like full game scaffolds, multi-scene architecture, physics-heavy gameplay, or performance audits. MUST use for any Phaser request: creating a game, loading assets, physics, sprites, input, tilemaps, collisions, scene management, Phaser.Game, Phaser.Scene, this.physics, this.anims, preload/create/update, or help with a Phaser game. Target Phaser 3.60+ unless Phaser 2/CE is explicitly requested.
development2
unity-editor
Write Unity Editor scripts — custom inspectors, editor windows, property drawers, scene handles, gizmos, and menu items. Auto-triages: Quick (single editor script — one inspector, one window, one drawer) or Deep (multi-file editor tooling — level editors, custom pipelines, inspector + gizmos + window combos). MUST use for ANY Unity Editor code request — custom inspectors, editor windows, property drawers, gizmos, handles, menu items, scene tools. Triggers: "custom inspector," "editor window," "property drawer," "gizmo," "scene handles," "menu item," "editor tool," "custom editor." Do not use for runtime scripts (unity-code), UI Toolkit (unity-uitoolkit), tests (unity-test-unit), or debugging (unity-debug).
tools2
unity-profiler
Analyze Unity Profiler data and generate severity-ranked performance reports. Covers CPU frame analysis, memory/GC pressure, rendering bottlenecks, and custom profiling instrumentation. Use when the user says "analyze the profiler data," "performance report," "why is my game lagging," "check for GC spikes," "frame drops," "draw call issues," "memory leak," or has completed a profiling session and needs interpretation. Also use when the user wants to add custom ProfilerMarkers, measure specific code sections, or instrument their code for profiling. Read-only analysis — suggests fixes but does not modify code (except Custom mode which generates profiling snippets).
development2
unity-prototype
Build playable Unity prototypes from game ideas — analyze the concept, scope an MVP, break into tasks, and delegate to subagents (unity-code, unity-uitoolkit) to deliver a working, testable prototype scene. Bridges the gap between planning and full implementation by producing running code fast. Use when the user says "prototype this," "build a quick prototype," "make an MVP," "proof of concept," "I have a game idea," "let me test this mechanic," "game jam," "hack this together," "rapid prototype," or describes a game concept and wants to see it running. Also use when the user wants to validate a mechanic before full implementation, or says "can we quickly build [concept]," "I want to try out [idea],\" or "let's see if this works." Do not use for production-quality features (unity-code), design documents (unity-spec), or planning without implementation (unity-plan).
tools2
unity-add-log
Generate formatted Debug.Log snippets for tracing values, events, or methods in Unity C# without modifying project files directly.
development2
unity-optimize
Optimize Unity project performance — code hot paths, build settings, rendering, memory, physics, and platform-specific tuning. Auto-triages: Code (GC elimination, allocation-free patterns, Jobs/Burst migration in specific scripts), Settings (Player Settings, Quality Settings, Physics Settings, platform config for target hardware), Audit (full project performance sweep with ranked findings and applied fixes). MUST use for ANY Unity optimization request — "make this faster," "reduce GC," "optimize for mobile," "reduce build size," "too many draw calls," "memory issues," "frame drops during gameplay," "performance pass," "optimize my project," "why is my game slow," even when the user doesn't say "optimize" explicitly but describes performance symptoms. Do not use for profiler data analysis (unity-profiler), code refactoring without perf focus (unity-code), or general code writing (unity-code).
development2
session-retrospective
Post-session retrospective engine — automatically triggered after a session or task completes. Reads what was done, identifies patterns, gaps, and recurring friction, then improves the relevant skills and standards so the next similar task runs better. Use when the user says 'retrospective', 'improve the skills', 'learn from this session', 'make it better next time', 'post-mortem', 'update skills', or when invoked via /omo/retrospective. Also triggered automatically after every session idle event. ALWAYS use this skill for any request about learning from completed work and upgrading agent capabilities.
data-ai2
ui-ux
UI/UX design intelligence. 67 styles, 96 palettes, 57 font pairings, 25 charts, 13 stacks (React, Next.js, Vue, Svelte, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui). Actions: plan, build, create, design, implement, review, fix, improve, optimize, enhance, refactor, check UI/UX code. Projects: website, landing page, dashboard, admin panel, e-commerce, SaaS, portfolio, blog, mobile app, .html, .tsx, .vue, .svelte. Elements: button, modal, navbar, sidebar, card, table, form, chart. Styles: glassmorphism, claymorphism, minimalism, brutalism, neumorphism, bento grid, dark mode, responsive, skeuomorphism, flat design. Topics: color palette, accessibility, animation, layout, typography, font pairing, spacing, hover, shadow, gradient. Integrations: shadcn/ui MCP for component search and examples.
tools2
unity-test-unit
Use this skill to write Edit Mode and Play Mode unit tests for Unity C# using Arrange-Act-Assert pattern with 10+ test cases per class. Use when the user says "write tests," "unit test this class," "add test coverage," "create tests for X," or wants automated verification of a MonoBehaviour, ScriptableObject, or utility class. Do not use for manual QA test case documentation — use unity-test-case for that.
tools2
visual-explainer
Generate beautiful, self-contained HTML pages that visually explain systems, code changes, plans, and data. Use when the user asks for a diagram, architecture overview, diff review, plan review, project recap, comparison table, or any visual explanation of technical concepts. Also use proactively when you are about to render a complex ASCII table (4+ rows or 3+ columns) — present it as a styled HTML page instead.
development2
unity-review-local
Use this skill to review locally changed Unity C# files. Use when the user asks to "review my code", "check changes", or specific file paths without PR context.
development2
Use this skill when the user needs to read, create, or review PDF files where rendering and layout matter. Covers generating PDFs programmatically (reportlab), extracting data from PDFs (pdfplumber, pypdf), and visually validating PDF output by rendering pages (Poppler). Use it even when the user just says "make this into a PDF," "extract data from this PDF," or "check if this PDF looks right." Do not use for simple text extraction where layout doesn't matter.
testing1
screenshot
Use this skill when the user explicitly asks for a desktop or system screenshot — full screen, a specific app or window, or a pixel region. Also use when tool-specific capture capabilities are unavailable and an OS-level capture is needed. Supports macOS (screencapture), Linux (scrot/gnome-screenshot), and Windows (PowerShell). Do not use for in-app screenshots or Unity scene captures — those have dedicated tools.
tools1
unity-review
Use this skill to review Unity code changes, GitHub pull requests, or audit entire Unity projects for quality. Covers three modes: local review (inline // REVIEW comments on changed C# files), PR review (parallel specialist reviews of .cs, .prefab, .shader files with APPROVE/REQUEST_CHANGES via GitHub API), and project audit (A-F grades on architecture, performance, best practices, tech debt as HTML report). MUST use whenever the user mentions reviewing code, checking changes, evaluating a pull request, auditing project quality, or assessing merge readiness. Also use when they say "look over my changes," "check this PR," "code review," "is this ready to merge," "rate this codebase," "tech debt," "quality report," "check the prefabs," "review the assets," or wants any feedback on Unity code, assets, or architecture before merging or committing.
development1
unity-test-case
Use this skill to generate structured QA test cases as an HTML document — covering happy paths, edge cases, boundary values, and negative tests. Use when the user needs a manual test plan, says "generate test cases," "QA test plan," "what should we test," or wants documented test scenarios before a release. Outputs to Documents/TestCases/. Do not use for automated unit tests — use unity-test-unit for that.
development1
unity-uitoolkit
Use this skill to build runtime UI with Unity UI Toolkit — UXML templates, USS styling, C# controller bindings, and custom controls. Accepts design docs, images, or verbal descriptions as input and asks clarifying questions for ambiguous requirements. Use when the user says "create a UI screen," "build a menu," "make a HUD," "UI from this mockup," or any request involving UXML, USS, or runtime UI panels. Do not use for editor-only UI — use unity-code for that.
tools1
flutter-debug
Unified Flutter/Dart debugging skill — diagnose and fix any Flutter bug, from Dart analysis errors to complex intermittent issues. Automatically triages into the right mode: Fix (Dart compile/analysis errors, auto-fix loop), Quick (runtime crashes, widget errors, interactive 2-3 solution proposals), Deep (intermittent/multi-system, read-only investigation report), or Log (generate structured logging snippets). Use when the user says "fix this bug," "fix this error," "why is this broken," "debug this," "root cause analysis," "auto fix," "resolve these analysis errors," "add debug logs," "trace this value," "something is wrong," "this isn't working," or describes unexpected behavior, pastes a stack trace, or reports "works on emulator but not on device." Also use when a bug has resisted 2+ fix attempts and needs documented analysis.
development1
spreadsheet
Use this skill when the user needs to create, edit, analyze, or format spreadsheet data — .xlsx, .csv, or .tsv files. Covers adding formulas, derived columns, charts, pivot analysis, data cleaning, formatting, and formula-aware recalculation. Use it when the user has tabular data and wants to explore, transform, or visualize it, even if they don't explicitly mention "spreadsheet" or "CSV." Also use when the user says "add a column," "summarize this data," or "make a chart from this file."
development1
sisyphus-work
Autonomous goal execution engine — scans Docs/Goals/**/*.md (recursively, including feature subfolders) for uncompleted goals and executes ALL of them without asking, stopping, or confirming. Checks Docs/Specs/ for related design specs before implementation and mandatorily updates specs after each goal completes via unity-spec. Thinks, decides, plans, delegates, and verifies entirely on its own. Creates task breakdowns, verifies with domain-specific checks (Unity console, build commands, static analysis), and performs a final goal review gate. Use when the user says 'execute goals,' 'run all goals,' 'autonomous mode,' 'sisyphus work,' 'just do everything,' 'do the goals,' 'start working,' 'execute the plan,' 'implement everything,' or invokes /omo/sisyphus-work. Also use when goal files exist in Docs/Goals/ and the user wants unattended execution. MUST use for any autonomous, no-questions-asked goal completion from goal documents.
development1
sisyphus-goal
Interactive goal creation and update skill — collaboratively defines structured goal files through clarifying questions and writes them to Docs/Goals/{feature-name}/{kebab-case-task}.md using the [Feature] Task title format. Every invocation MUST produce or update a goal document on disk. Always scans Docs/Goals/ and all subfolders to avoid duplicates and groups related goals by feature folder. Use when the user wants to create, update, or revise goals — 'new goal,' 'add a goal,' 'create a goal,' 'I want to achieve X,' 'plan this as a goal,' 'break this down into goals,' 'define acceptance criteria,' 'change the priority,' 'add a criterion,' 'bump priority.' Also triggers on 'I want to build X,' 'capture this as an executable plan,' 'write acceptance criteria,' 'scope this feature,' 'plan this,' 'spec this out,' or any work planned with verifiable criteria before coding starts. Produces one goal per file that sisyphus-work executes autonomously.
development1
git-squash
Use this skill to squash a branch's commits into clean, logical units before merging a PR. Use it when the user wants to "clean up commits," "squash before merge," "combine WIP commits," or "make the history cleaner." Analyzes commit history, groups related changes, presents a squash plan for approval, then executes via interactive rebase. Do not use for amending a single commit — use git-comment for that.
documentation1
sisyphus-improve
Quality refinement engine — reads Docs/Goals/**/*.md (recursively, including feature subfolders), assesses work output against acceptance criteria, identifies gaps, delegates targeted improvements, and verifies results. Use after sisyphus-work completes, when the user says 'improve this,' 'make it better,' 'check against goals,' 'refine the work,' 'quality pass,' 'sisyphus improve,' or wants post-execution quality review. Runs autonomously like sisyphus-work but focused on QUALITY over COMPLETION.
development1
flutter-ui
Unified Flutter UI skill — compose screens, widgets, themes, and responsive layouts. Auto-triages: Quick (single widget, one screen, layout fix, simple styling) or Deep (multi-screen systems, custom themes, responsive breakpoints, animations, component libraries). MUST use for ANY Flutter UI composition request — building screens, styling widgets, theming, responsive layout, design systems, onboarding flows. Triggers: "build a screen," "create a widget," "add dark mode," "responsive layout," "fix padding," "design system," "component library," "onboarding flow." Do not use for business logic, state management, or services (flutter-code), tests (flutter-test), or debugging (flutter-debug).
development1
flutter-profiler
Analyze Flutter performance data and generate severity-ranked reports. Covers CPU frame analysis, memory/GC pressure, frame rendering and jank, and custom profiling snippets. Use when the user says "analyze profiler data," "performance report," "why is my app slow," "check for jank," "memory leak," "GC spikes," "frame drops," or has completed a DevTools profiling session and needs interpretation. Also use when the user wants to add custom performance markers, timeline events, or profiling instrumentation to their Flutter code. Read-only analysis — suggests fixes but does not modify code.
tools1
nextjs-backend
Unified Next.js backend coding skill — write API routes, server actions, database queries, authentication, and multi-tenant logic using the App Router. Auto-triages: Quick (single route handler, one server action, simple CRUD endpoint, auth check) or Deep (multi-file features: full auth system, multi-tenant setup, CRUD with pagination + validation, webhook integrations). MUST use for ANY Next.js backend request — API routes, route handlers, server actions, database queries (Prisma/Drizzle), JWT auth, middleware, multi-tenant architecture, webhook processing, or server-side data mutations. Triggers: "create an API route," "add a server action," "connect to PostgreSQL," "add authentication," "protect this route," "multi-tenant," "webhook handler," "database query," "Prisma schema," "Drizzle query," "JWT," "middleware auth." Do not use for frontend React components, styling, or client-side state management.
tools1
bash-check
Use this skill to validate a shell script before committing or deploying — runs bash -n syntax checks and ShellCheck static analysis, then produces a findings report. Use it whenever the user has a .sh or .bash script they want checked for bugs, unsafe patterns, or POSIX compliance, even if they just say "is this script OK" or "check this for issues." Do not use for refactoring or optimizing scripts — use bash-optimize for that.
development1
bash-install
Use this skill when the user needs to install a CLI tool, library, runtime, or dependency on their machine or CI environment — handles package manager selection, auto-retry with fallback strategies, and post-install verification. Use it even when the user just says "I need X set up" or "get me Y installed" without specifying how. Do not use for npm/pip install of project dependencies managed by a package file — those are standard tool calls.
tools1
bash-optimize
Use this skill to refactor an existing shell script for clarity, performance, and maintainability without changing its behavior. Use it when the user says a script is "messy," "hard to read," "slow," or asks to "clean up" or "improve" a .sh file. Covers removing unnecessary subshells, simplifying conditionals, improving variable quoting, and applying best practices. Do not use for validation-only tasks — use bash-check for that.
development1
database-design
Unified PostgreSQL database design skill — schema creation, migrations, indexing, multi-tenancy, connection pooling, seed data, and ORM integration (Prisma/Drizzle). Auto-triages: Quick (single table, add column, create index, simple migration) or Deep (multi-table schema, multi-tenant setup, migration strategy, full data model). MUST use for ANY PostgreSQL schema or database design request. Triggers: "design a schema," "create a table," "Prisma model," "Drizzle schema," "add an index," "migration strategy," "multi-tenant database," "normalize this," "PostgreSQL schema," "foreign key," "RLS policy," "JSONB vs relational," "cursor pagination," "soft deletes," "audit columns," "seed data," "connection pooling," "PgBouncer," "pool sizing," "factory pattern." Do not use for application-layer query optimization (nextjs-backend) or database administration (backups, replication).
development1
flatbuffers-coder
Use this skill when the user needs to work with FlatBuffers in a Unity project — defining .fbs schemas, generating C# code via flatc, or implementing binary serialization/deserialization. Use it whenever the user mentions FlatBuffers, .fbs files, flatc, or needs zero-copy binary serialization for game data, network packets, or save files, even if they just say "I need fast binary serialization" without naming FlatBuffers specifically.
development1
flutter-code
Unified Flutter/Dart coding skill — write, extend, or optimize Dart code across all domains. Auto-triages: Quick (single-file widget, provider, model, utility, service), Deep (multi-file features, state machines, system builds, refactors spanning 3+ files), Optimize (simplify/clean up without behavior change). MUST use for ANY Flutter/Dart code request — writing widgets, creating providers, implementing features, building services, refactoring, cleaning up code, reducing rebuilds. Triggers: "write a widget," "create a provider," "build a feature," "implement service," "refactor," "clean up," "simplify," "optimize this." Do not use for UI-focused layout work (flutter-ui), tests (flutter-test), or debugging (flutter-debug).
tools1
flutter-review
Unified Flutter review skill — reviews local changes, GitHub PRs, or full Flutter projects. For local changes, adds inline // REVIEW comments to .dart files. For GitHub PRs, classifies changed files by type (.dart, .yaml, assets), assesses change size, spawns parallel specialist reviews (code quality, architecture, state management, performance), aggregates findings, and posts APPROVE or REQUEST_CHANGES via GitHub API. For project audits, grades architecture, code style, performance, best practices A-F with an HTML report. Use whenever the user says "review my code," "review this PR," "check my changes," "review PR #123," "audit this project," "code quality report," "is this ready to merge," or wants any form of Flutter code or architecture review. Also triggers on "check the widgets," "review state management," or "review the providers."
development1
flutter-test
Write Flutter tests with auto-triage across Unit, Widget, and Integration modes. Uses AAA (Arrange-Act-Assert) pattern, mocktail for mocking, ProviderContainer for Riverpod providers, and minimum 10 test cases per class. Use when the user says "write tests," "unit test," "widget test," "integration test," "add test coverage," "test this feature," "test this provider," "test this screen," or wants automated verification of any Dart class, widget, or user flow. Do not use for manual QA test plans — use a test-case skill for that.
testing1
git-clear
Use this skill to delete all comments from a GitHub Pull Request — both issue/timeline comments and inline review comments. Use it when the user wants to clear stale AI review comments, remove bot-generated noise, or start a fresh review cycle on a PR. Also use when the user says "clean up this PR" or "remove old comments" in the context of a GitHub PR.
documentation1
git-comment
Use this skill to rewrite the most recent local commit message to be clean, imperative, and well-structured. Use it when the last commit message is unclear, auto-generated, or the user says "fix my commit message," "that commit message is bad," or "reword the last commit." Never touches pushed commits. Do not use for making new commits — use git-commit for that.
documentation1
git-commit
Use this skill to stage and commit changes with a clean, imperative-mood commit message. Use it whenever the user wants to save their work, says "commit this," "save my changes," or "commit what I've done." Handles both staged and unstaged files, writes structured messages with bullet points for multi-change commits. Never pushes to remote. Do not use for amending — use git-comment for that.
documentation1
git-description
Use this skill to generate and apply a structured description for a GitHub Pull Request by deep-investigating all changed files. Use it when the user opens a new PR and needs a description, when a PR's description is missing or stale, or when the user says "describe this PR" or "write a summary for this PR." Applies the result directly via gh pr edit.
documentation1
mermaid
Use this skill to create Mermaid diagrams — flowcharts, sequence diagrams, state machines, class diagrams, and architecture diagrams. Use it when the user wants to visualize a system, document a flow, explain an architecture, or create any kind of technical diagram, even if they don't mention "Mermaid" specifically. Also use when the user says "draw this," "show me how X flows," or "visualize the architecture."
testing1