Firzus
UserAgent Skills for AI coding assistants
Categories
Indexed Skills (72)
implement
Deliver prepared work with test-first verification, current documentation, and a verified handoff.
improve-architecture
Find evidenced architecture improvements, present recommendations, and delegate each selected refactor through verified delivery and report updates.
interview
Prepare approved Linear work briefs through a continuous interview, with progressive framing for large goals.
setup-codex
Configure a reviewed Codex Operating Policy through model_instructions_file and offer low model verbosity with separate consent, backups, and verification.
unslop
Audit a codebase for dead code and AI slop, clear each suspect against the legitimate reason it exists, and remove only what the evidence carries.
imagegen
Generates or edits raster images with AI (photos, illustrations, textures, sprites, mockups, logos, infographics), including transparent-background cutouts. Use when the task needs an AI-created or AI-edited bitmap; not for SVG/vector or code-native visuals.
unity
Picks the Unity tool for the job and applies it. Routes each need to one chosen default — UI Toolkit, Awaitable, Addressables, Input System, URP, Netcode for GameObjects, Multiplayer Services, Dedicated Server, Unity CLI — and keeps the project CoreCLR-ready for the Mono removal in 6.8. Use when writing, reviewing, or architecting Unity code, when building a custom Editor tool or inspector, when driving the Unity Editor from an agent, when choosing between two Unity tools that do the same job, or when the user mentions Unity, UI Toolkit, UGUI, Addressables, URP, HDRP, DOTS, ECS, netcode, custom inspector, SerializeReference, CoreCLR, IL2CPP, Unity CLI, Unity MCP, or Unity build and performance questions.
design-system
Create a project's design system — design read, dials, tokens, themes — recorded in DESIGN.md. Use when the user wants a design system, theme, or design tokens for a site or app, or when the frontend-design pipeline reaches phase 1.
tauri
Tauri 2 development and validation. Use for src-tauri, Rust/frontend IPC, capabilities and plugins, windows, mobile entry points, packaging, or inspecting a running Tauri app with tauri-agent-kit.
canvas
Renders a standalone analytical artifact — architecture review, audit, quantitative analysis, data-heavy report, comparison, timeline, progress dashboard — as a Markdown-backed reader or standalone design HTML beside the chat. Works in any environment (Codex, Claude Code, Cursor, generic agents). Use when the structured output IS the deliverable and benefits from visual layout, and whenever updating an existing canvas file.
code-review
Review a scoped diff against requirements, project standards, and correctness using independent read-only reviewers. Use for branch, PR, or uncommitted-change review, and implementation review checkpoints.
debug
Investigate bugs, failing tests, crashes, unexpected behavior, and performance regressions through reproducible evidence and hypothesis testing. Use when asked to debug or diagnose; apply and verify a fix only when authorized.
deep-research
Investigate a question across sources with delegated research, adversarial verification, and a reusable evidence dossier. Use for explicit deep research, substantial comparisons, evidence accumulation, or updating a research dossier.
prototype
Resolve a bounded design or feasibility question through an observable experiment. Use for authorized interface, logic, or technical exploration, including prepared Prototype issues; not settled implementation or source-only research.
security-review
Perform source-only security audits of codebases, APIs, services, and applications. Use for security questions, focused vulnerability reviews, or explicitly requested full audits.
skills
Install, update, remove and author Agent Skills with the `skills` CLI (skills.sh, vercel-labs/skills). Use when adding a skill from a repo, sharing skills across agents, choosing project vs global scope, publishing a skill repo, or debugging a skill the agent never fires.
video-report
Record video evidence with FFmpeg after a visual bug fix or when asked for a recorded demo of a web page, native application, or game.
writing-for-agents
Create, update, and review AGENTS.md files, skills, and their references. Propose project-instruction changes for approval before writing.
youtube-transcript
Transcribes YouTube videos and produces evidence-backed learning reports. Use when asked to extract YouTube captions, transcribe a video's audio, or explain what a video teaches using speech and on-screen demonstrations.
design-references
Find inspected UI references for visual inspiration or interaction decisions, especially in editors and professional tools. Also use for explicit maintenance of the bundled design-reference catalog.
figma-to-code
Implements a Figma design from a Figma URL, using the Figma MCP server: extracts design context, tokens (colors, typography, spacing), and assets (SVG, PNG), builds the UI in the project's stack — or as a standalone page for design review — then iterates until the render matches the node geometry Figma reports, to the pixel, and the reference screenshot. Use when the user provides a Figma link and wants it implemented, asks to "implement this Figma design", "build this screen from Figma", "make it match the Figma", or mentions figma-to-code.
nextjs
Next.js 16+ App Router with React Server Components and TypeScript. Use when working in app/, layouts, pages, route handlers, proxy.ts, Server Actions, "use client"/"use server"/"use cache" boundaries, caching and revalidation, typed routes, view transitions, Cache Components adoption, instant navigation optimization, runtime verification, or Partial Prefetching adoption.
payload-cms
Payload CMS 3.x, the Next.js-native TypeScript headless CMS. Use when working in payload.config.ts, collections, fields, access control, hooks, the Local API, versions and drafts, database or storage adapters, plugin authoring, or the official MCP plugin.
tanstack-query
TanStack Query v5 (React Query) async state management for React. Use when implementing or reviewing code that uses `@tanstack/react-query` — query keys, queryOptions, useQuery/useSuspenseQuery, mutations and invalidation, QueryClient defaults, SSR hydration, or infinite queries.
tanstack-router
TanStack Router v1 for React — file-based routing, type-safe navigation, validated search params, loaders and router context. Use when working in a project that uses `@tanstack/react-router` — routeTree.gen.ts, createFileRoute, Link/useNavigate, validateSearch, loaderDeps, route masking, or code splitting.
tanstack-start
TanStack Start, the full-stack React framework built on TanStack Router with Vite or Rsbuild. Use when setting up or reviewing a `@tanstack/react-start` project — the build-tool plugin, SSR and selective SSR, server functions (`createServerFn`), server routes, middleware, prerendering, SPA mode, environment variables, or deployment targets.
tanstack-store
TanStack Store client-state management. Use when implementing or reviewing @tanstack/store or its framework adapters, immutable updates, derived stores, selectors, subscriptions, batching, scoped SSR state, or migrations from legacy Store/Derived/Effect APIs.
vite-plus
Vite+ projects using vite-plus and the vp CLI. Use when setting up or migrating to Vite+, configuring its checks, tests or library builds, debugging built-in versus script execution, or configuring workspace tasks and caching. Applies to Vite+ integrations, not every standalone Vite project.
zod
Zod schema validation. Use when writing or reviewing Zod schemas, parsing untrusted data, handling coercion, refinements, transforms or validation errors, integrating forms or APIs, exporting JSON Schema, or migrating Zod versions.
agent-browser
Drive a real browser from the CLI — navigate, snapshot, click, fill, screenshot, emulate. Use when the user wants to interact with a website, test or review a web UI, take screenshots, extract page data, or attach to a CDP endpoint (Chrome, Tauri/WebView2, Electron), or when another skill needs browser evidence.
artefact
Render a concept as a single self-contained HTML document (Tailwind + Mermaid via CDN), written to the OS temp directory and opened in the browser. Use when the user wants to visualize a concept, comparison, plan, or design (design tokens, an architecture, a data flow), or when another skill needs to present a visual document.
babysitting-pr
Monitor an open GitHub pull request until it is merge-ready by checking CI, diagnosing and fixing branch-related failures, addressing actionable review feedback, resolving safe merge conflicts, pushing fixes, and re-checking the result. Use when the user asks to babysit, watch, monitor, or keep an open PR green and ready to merge.
gamification
Gamification design for products, apps, and game meta-layers: design an engagement, habit, or retention system; choose or audit mechanics (points, badges, leaderboards, streaks, progress bars, variable rewards); diagnose mechanics that backfire (novelty decay, demotivated users, streak churn); or run an ethics / dark-pattern check.
reverse-engineer
Reverse engineer how an application implements a mechanism so it can be replicated. Use when the user asks how an app does something, how a feature or protocol works under the hood, or wants design notes to reimplement a behavior (download chunking, retry, sync, auth flow, caching) from a repository, an installed build, or a closed-source app known only by name.
simplify
Simplify and refine recently modified code for clarity and consistency. Use after writing code to improve readability without changing functionality.
software-architecture
Stack-agnostic software architecture guidance for any kind of software — game, desktop app, web/SPA, backend service, CLI. Picks the lightest macro structure and micro/runtime pattern for the problem, with per-domain examples, costs, and explicit "avoid when" guidance. Use when designing or refactoring an app's architecture, choosing how to layer or modularize code, drawing module/process boundaries (including IPC), managing dependencies and coupling, structuring state and persistence, handling errors/logging/config/async across boundaries, or when the user mentions layering, hexagonal/clean/onion, ports and adapters, dependency inversion, DDD, ECS, design patterns, state management, CQRS, or over-engineering.
adaptive-audio
Architecture blueprint for game audio systems: adaptive music, mix states, voice management, spatial audio, DSP/synthesis, middleware choices, audio accessibility, and production pipelines. Use when designing or building game music, ambience, occlusion, procedural audio, sound concurrency, or when music cuts mid-beat, layers drift, ducking pumps, dialogue is stolen, the audio thread glitches, or accessibility is missing.
camera-system
Architecture blueprint for game cameras across genres: virtual-camera stacks, blending brains, orbit/follow rigs, collision and occlusion, combat lock-on, genre-specific cameras, damping math, shake, FOV, photo mode, and comfort. Use when designing or building game cameras, dynamic framing, camera collision, lock-on, shake, split-screen, VR comfort, or when the camera jitters, fights the player, or causes motion sickness.
character-controller
Design, build, or diagnose Unreal Engine third-person action-RPG character controllers with Mover and Network Prediction. Use for C++ locomotion architecture, grounded/falling/swimming/climbing/gliding modes, moving bases, teleport, dedicated-server rollback, combat displacement, Motion Warping, animation handoff, jitter, correction storms, or movement desync. This skill is Mover-first and does not fall back to Character Movement Component.
cinematic-system
Architecture blueprint for game cutscene systems: timeline tracks, bindings, realtime vs video playback, gameplay-to-cutscene transitions, skip/replay guarantees, cinematography craft, production pipeline, QTEs, branching and playable cinematics, virtual production, and accessibility. Use when designing cutscenes, timelines, in-game cinematics, video playback, skip systems, QTEs, or when skips break quest state, bindings T-pose, assets stream late, or film grammar feels off.
combat-system
Architecture blueprint for melee, ranged, RPG, and turn-based combat systems: attack graphs, hit detection, damage pipelines, stagger/poise, defensive kits, gunplay, aim assist, lag compensation, status effects, formulas, balance, and combat feel. Use when designing combat, combos, hitboxes, gunplay, dodge/parry, stagger, status effects, TTK/DPS/EHP balance, or when combat feels mushy, shots miss, or damage numbers run away.
coop-session
Architecture blueprint for drop-in/drop-out co-op and multiplayer sessions: lobby lifecycle, join rules, authority, replication, late-join snapshots, anti-grief content rules, scaling, topology, matchmaking, host migration, split-screen, loot distribution, revive, cross-play, and co-op design. Use when designing co-op, lobbies, netcode, matchmaking, or when late joiners desync, hosts gain unfair advantage, rewards conflict, or co-op is not fun.
dialogue-system
Architecture blueprint for dialogue systems in open-world, RPG, and narrative games: graph data, separated text/conditions, runtime sessions, barks, branching patterns, skill checks, presentation, subtitles, VO, localization, lip-sync, and generative NPC constraints. Use when designing conversations, dialogue trees, barks, subtitles, VO pipelines, or when players get stuck in dialogue, text overflows, NPCs freeze, or dialogue wheels mislead.
enemy-ai-framework
Architecture blueprint for game AI across genres: FSM/HSM, behavior trees, GOAP/HTN/utility AI, brain-intent-execution separation, perception, aggro, threat selection, attack tokens, lifecycle, AI LoD, group coordination, pathfinding, avoidance, crowds, and believability. Use when designing enemy AI, behavior trees, spawning, stealth/RTS/racing/sim AI, or when enemies swarm, flip targets, path badly, or feel robotic.
figma-to-unity
Implements Figma designs (frames, components, HUD elements, menus) as Unity UI Toolkit interfaces — UXML hierarchy, USS styles mapped to design tokens, exported sprite assets, and a minimal C# controller — using the Figma MCP server (get_design_context, get_metadata, get_screenshot, get_variable_defs). Use when the user provides a Figma URL and wants it implemented in a Unity project, asks to "implement this Figma design in Unity", "build this HUD/menu/component from Figma", "convert Figma to UXML/USS", or mentions figma-to-unity.
hud-system
Architecture blueprint for in-game HUD systems: event-driven read-only UI, layout hierarchy, dynamic visibility, damage numbers, bars, gauges, notifications, prompts, quest trackers, diegetic/spatial HUDs, accessibility, nameplates, reticles, and world-space indicators. Use when designing HUDs, health bars, prompts, damage text, accessible UI, or when HUDs cost frame time, break on TVs/ultrawide, or fail colorblind/low-vision players.
loot-drop-system
Architecture blueprint for loot tables, world drops, respawn, claim gating, multiplayer loot, drop perception, and loot-box compliance across open-world, ARPG, looter-shooter, and MMO games. Covers weighted tables, guaranteed drops, magic find, scatter/despawn, personal/FFA loot, pseudo-randomness, odds disclosure, and anti-cheat. Use when designing drops, chests, nodes, respawn, drop rates, or when loot despawns, RNG feels rigged, rewards conflict, or odds need disclosure.
magica-cloth-2
Author and tune cloth physics with Magica Cloth 2 in Unity, code-first via the Unity MCP (Unity_RunCommand) — BoneCloth / MeshCloth / BoneSpring setup, capsule/ sphere/plane colliders, constraint parameters, runtime build, and in-editor visual testing. Use when adding or fixing cloth/jiggle on a character (skirt, hair, cape, tail, chest, accessories), when cloth clips through the body, jitters, feels too stiff/floppy, penetrates on fast motion, or when the user mentions Magica Cloth, MagicaCloth, cloth simulation, or dynamic bones in a Unity project.
menu-ui-manager
Architecture blueprint for game menus and screen frameworks: central routers, layered screen stacks, declarative contracts, hub navigation, multi-input focus, pause, transitions, settings, rebinding, modal APIs, UI architecture, performance, accessibility, and menu feel. Use when designing menus, pause screens, settings, UI navigation, screen stacks, or when gamepad focus dies, back buttons misbehave, menus leak input, or UI hurts frame rate.
minimap-worldmap
Architecture blueprint for minimap and full-screen world map systems: content baking, tiled zoom pyramids, world-to-map transforms, shared marker registries, clustering, label collision, fog of war, multi-layer maps, pan/zoom/pins, fast travel, breadcrumbs, tile rendering, and cross-genre UX. Use when designing minimaps, world maps, markers, fog reveal, diegetic compasses, or when markers drift, fog reverts, precision breaks, or maps hurt performance.
mount-system
Architecture and implementation workflow for player-controlled animal and creature mount systems in Unreal Engine. Design, build, refactor, or diagnose summon and dismiss, mount and safe-dismount transactions, possession and rider binding, Mover-based ground/flying/aquatic locomotion, collision, animation, camera, combat, replication, join-in-progress, reconnect, persistence, and cosmetic separation. Use for unsafe dismounts, competing displacement writers, mount correction storms, lifecycle desync, or gameplay-affecting skins. Route player transformations, generic traversal verbs, and mechanical multi-crew vehicles to their dedicated systems.
open-world-streaming
Architecture blueprint for open-world streaming: world partitioning, cells, streaming sources, async load/unload, hysteresis, layered streaming, memory and frame budgets, HLOD, fast-travel gates, virtualized geometry/textures, large-world rendering, procedural generation, and living-world simulation. Use when designing open worlds, level streaming, chunk loading, seamless worlds, procedural terrain, or diagnosing hitches, pop-in, memory issues, pool thrash, or non-determinism.
progression-economy
Systems architecture blueprint for game progression and economy: stat curves, XP families, skill trees, soft caps, stat aggregation, wallets, faucets/sinks, energy gates, battle passes, seasons, server-authoritative transactions, idempotency, and int64 currencies. Use when designing leveling, XP curves, talents, prestige, currencies, economy balance, stamina, reward grants, or when balances drift, grants duplicate, inflation appears, or curve edits break saves.
quest-system
Architecture blueprint for quest systems in open-world, RPG, and live-service games: quest/step/objective data, conditions/actions, prerequisite DAGs, event-driven runtime, world changes, shared-NPC conflicts, procedural quests, scripting, save migration, debug tools, telemetry, journals, markers, and authoring workflows. Use when designing quests, missions, objectives, logs, markers, commissions, branching quests, or when chains break, flags sprawl, saves corrupt, or players cannot find objectives.
save-persistence
Architecture blueprint for game save and persistence systems: versioned stores, stable IDs, world-state deltas, checkpoint/save-anywhere/continuous/server saves, serialization formats, schema evolution, compression, atomic writes, async saves, MMO persistence, idempotency, ledgers, rollback, slots, autosave, cloud saves, cross-progression, and tamper protection. Use when designing save/load, autosave, world persistence, MMO player data, or when saves corrupt, bloat, break after updates, or duplicate on retry.
scene-flow-manager
Architecture blueprint for game application flow: context state machines, boot/title/login/world/cinematic states, additive scene composition, atomic transitions, loading screens, online auth and enter-world handshakes, queues, reconnection, soft reset, async loading, lifecycle, suspend/resume, patching, FTUE, and loading UX. Use when designing boot flows, scene loading, login, transitions, onboarding, PSO warmup, or when relogin differs, traversal stutters, queues fail, or resume crashes.
stylized-rendering
Architecture blueprint for stylized / non-photoreal (NPR) cel-shaded game rendering in anime and painterly lineages. Covers ramp/step lighting, channel-packed control maps, tinted shadows, outline techniques, SDF face shadows, hair/eye shading, MatCap, rim light, tonemapping, bloom, art-direction calibration, and Unity 6 / UE5 mappings. Use when building toon, cel, anime, or stylized looks, NPR shaders, outlines, ramp lighting, face shadows, or when characters read flat, look uncanny, or outlines break.
teleport-map-unlock
Architecture blueprint for fast travel, waypoint networks, and map unlocking: region reveal, fog-of-war data, tower/statue/item/proximity unlocks, waypoint registries, stable IDs, spawn placement, discovery states, atomic teleport sequences, streaming gates, camera warps, large-world precision, networked teleport, and travel policy. Use when designing fast travel, teleports, waypoint activation, map reveal, or when players fall through the world, cameras whiplash, maps drown in icons, or quests leave travel locked.
traversal-system
Architecture blueprint for traversal above the character controller: traversability markup, surface probing, volumes, anchors, climb/glide/swim, grapple, parkour, wall-run, web-swinging, mantle/vault traces, motion warping, IK, vehicles, stamina, networking, readability, and design constraints. Use when designing climbing, parkour, gliding, swimming, vehicles, grappling, traversal progression, or when players escape bounds, probes fail, automation overrides intent, or traversal trivializes the world.
ue5-aaa-best-practices
Senior Unreal Engine 5 developer guidance: DO/DON'T best practices for building production-quality (AAA-level) games with UE5 (5.4/5.5/5.6+). Covers the C++ foundation / Blueprint leaf doctrine, Gameplay Ability System, Subsystems, GameFeatures, CommonUI and MVVM Viewmodel, Enhanced Input, MetaSounds, soft references and Asset Manager, World Partition, Nanite/Lumen decisions, tick discipline, Perforce/DDC/Horde CI, and the automation/Gauntlet test stack. Use when developing, reviewing, or architecting an Unreal Engine project, or when the user mentions Unreal, UE5, Blueprints, UMG, GAS, Nanite, Lumen, World Partition, or Unreal-specific performance and build questions.
world-time-weather
Architecture blueprint for time-of-day, weather, seasons, and climate systems: central game clocks, time scale, calendars, server/diegetic clocks, weather profiles, regional schedules, overrides, volumetric clouds, storms, systemic weather effects, survival temperature, event schedulers, daily resets, TOD lighting, rendering integration, and persistence. Use when designing day/night, weather, seasons, dynamic skies, hazards, resets, or when weather pops, forced weather leaks, the sun jitters, or time-skip floods events.
adsense
AdSense publisher monetization — improve ad revenue on a website without risking the account. Use when the user wants to raise AdSense revenue or RPM, place ad units or choose between Auto ads and manual units, diagnose an earnings, coverage, or ads-not-showing problem, integrate or fix ad code in a React/Next.js or SPA codebase, clear an AdSense policy, consent/CMP, ads.txt, or account-approval blocker, reduce ad-caused CLS/LCP damage, or judge a move to Ad Manager or a managed ad network.
astryx
Astryx (@astryxdesign) — Meta's open-source, agent-ready React + StyleX design system. Covers setup (npm packages, npx astryx init, StyleX compiler), the agent discovery loop (astryx template/component/docs, MCP server, --dense), per-component subpath imports, the xstyle styling order, design tokens, the Theme provider and defineTheme theming, cascade-layer CSS wiring, Tailwind and library interop, layout primitives, and incremental migration. Use when the user mentions Astryx, @astryxdesign, npx astryx, xstyle, defineTheme, or wants to set up, build UI with, theme, review, or migrate to the Astryx design system.
dokploy-best-practices
Dokploy deployment and operations best practices. Use when the user deploys apps, databases, or Docker Compose stacks with Dokploy; configures domains, Traefik, Swarm health checks, CI/CD, backups, or webhooks; or troubleshoots Bad Gateway, 404, volume, database, or hardening issues.
extract-theme
Extracts the visual theme (colors, typography, radius, spacing, shadows) of a public website provided by the user and produces a shadcn/ui + Tailwind CSS v4 compatible token block, ready to paste into the project's `globals.css` (or `app.css`) inside `@theme` and `.dark`. Uses the `agent-browser` CLI to drive a real Chromium instance. Use when the user asks to "extract the theme of <url>", "reverse-engineer the design tokens of <site>", "copy the look of <site> into shadcn", "build a tailwind v4 theme from <url>", or mentions extract-theme / theme extraction / design tokens from a URL.
greyboxing
Greybox a site's pages — layout, UX, motion on placeholder content — from an existing DESIGN.md. Use when the user wants to greybox, lay out, or build the pages of a site whose design system exists, or when the frontend-design pipeline reaches phase 2.
real-content
Replace a greyboxed site's placeholders with real content — copy, imagery, data. Use when the user wants to fill a site with final content, write its copy, or source its imagery, or when the frontend-design pipeline reaches phase 3.
shaders
Guides building GPU-accelerated visual effects in React / Next.js with the `shaders` npm package from shaders.com — declarative `<Shader>` component trees, composition (stacking, nesting, blend modes, masking), reactive props, dynamic prop drivers (`auto-animate`, `mouse-position`, `map`), shape (SDF) effects, color space, performance budget (RTT / generator vs filter), and SSR safety. Use when the user asks to add a "shader background", "WebGPU effect", "aurora / plasma / swirl / glass / cursor-trail" hero, mentions shaders.com, the `shaders/react` package, the `<Shader>` component, or wants animated GPU-rendered visuals in a React or Next.js app. Not for raw GLSL / WebGL / Three.js / react-three-fiber work — those use different APIs.
swr
SWR v2 (stale-while-revalidate) data fetching for React and Next.js. Use when implementing or reviewing code that uses the `swr` package — useSWR hooks, cache keys, revalidation, mutations, pagination, or SWRConfig.
vite-plus-best-practices
Vite+ (vp) unified toolchain: unified config, built-in ≠ script, vp check, pack vs build, and migrate. Covers setup (vp install, env, managed runtime), the vp command surface (dev, build, check, test, run, pack, node), vite.config.ts blocks (including check), monorepo overrides, task caching, commit hooks, library packaging, and migrating Vite/Vitest/ESLint/Prettier projects. Use when the user mentions Vite+, vite-plus, the `vp` or `vpx` CLI, Oxlint/Oxfmt in a Vite context, tsdown, Vite Task, or asks to set up, configure, migrate, scaffold, review, or upgrade a Vite+ project.
web-extension
Design, build, port, debug, test, or prepare browser extensions for Chromium, Firefox, and Safari using the WebExtensions model. Use when work involves manifest.json, Manifest V3, extension service workers or background scripts, content scripts, extension messaging, browser permissions, toolbar actions, options pages, sidebars, cross-browser extension compatibility, store-ready packages, Chrome Web Store, Firefox Add-ons, or Safari Web Extensions. Also use when migrating an existing extension between browsers.
frontend-design
Design a site's whole frontend — design system, greyboxing, real content. Use when the user wants to design or redesign a site's frontend, or to build all the pages of a new site.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.