jeo-skills

Operate the Tons of Skills marketplace via the ccpi CLI and Claude plugin marketplace commands. Use for searching installing listing and updating Claude Code plugin packs with reversible rollbacks.

Run a reproduce → isolate → verify debugging workflow for concrete bugs, regressions, flaky failures, and environment-specific behavior. Use when the user already has a failing command, test, request, UI flow, or narrowed symptom and needs root-cause diagnosis or fix verification rather than raw log-line selection, broad test-policy design, PR review, or generic performance tuning.

INVOKE THIS SKILL when building ANY Deep Agents application. Covers create_deep_agent(), harness architecture, SKILL.md format, and configuration options.

INVOKE THIS SKILL when your Deep Agent needs memory, persistence, or filesystem access. Covers StateBackend (ephemeral), StoreBackend (persistent), FilesystemMiddleware, and CompositeBackend for routing.

INVOKE THIS SKILL when using subagents, task planning, or human approval in Deep Agents. Covers SubAgentMiddleware, TodoList for planning, and HITL interrupts.

Cross-runtime 2-stage pipeline for Claude Code, Codex/OMX, and Gemini/Antigravity/OMA: trace causal hypotheses, inject evidence into deep-interview style requirements crystallization, then hand off to the right runtime planner/executor.

Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Claude Code.

Agent-to-Agent horizontal messaging — Linux Foundation AAIF 2026 standard. Use when multiple agents across systems/vendors need to exchange tasks, results, or capabilities without sharing a runtime.

Analyze A/B test results with statistical significance, sample size validation, confidence intervals, and ship/extend/stop recommendations. Use when evaluating experiment results, checking if a test reached significance, interpreting split test data, or deciding whether to ship a variant.

Full academic research pipeline from discovery to publication — 4 reference pipelines, 27 modes, 39-agent ensemble (deep-research 8 modes, academic-paper 11 modes, academic-paper-reviewer 6 modes, academic-pipeline 10-stage orchestrator). Routes each request to the right pipeline and mode. Human-in-the-loop throughout. Plugin (upstream): claude plugin marketplace add Imbad0202/academic-research-skills

Map agent capabilities to standard benchmarks (SWE-bench, WebArena, OSWorld, GAIA, TauBench) plus custom regression packs. Use when you need defensible numbers on agent quality, not just vibes — before shipping, after every model change, and for vendor comparisons.

Layered agent safety — prompt-injection defense, data exfiltration prevention, tool-misuse blocking via NeMo Guardrails + Lakera / LLM Guard / Llama Guard. Use before any agent reads untrusted input or executes tools with real-world side effects.

tmux + Python agent lifecycle manager — start, stop, monitor, and assign tasks to AI agents running in tmux sessions. Use when managing multiple AI agents without a server, scheduling agent jobs via cron, monitoring agent heartbeats, or injecting skills into agent sessions. Zero dependencies beyond tmux + Python. Triggers on: "start agent", "stop agent", "monitor agent", "assign task to agent", "agent lifecycle", "agent scheduling", "tmux agent", "manage agents", "agent-manager".

Design tiered agent memory — core / recall / archival — with backend selection (mem0, Letta, Zep/Graphiti, LangMem) and temporal vs snapshot tradeoffs. Use when an agent needs persistent context across sessions, personalization, or temporal knowledge graphs.

Instrument LLM agents with traces, metrics, and replay. Use when an agent in production is silently failing, regressing, drifting, or burning tokens. Selects LangSmith / Langfuse / Phoenix, defines node-level spans, attaches evals to traces, and enables session replay.

Core principles for collaborative development with AI agents. Defines divide-and-conquer, context management, abstraction-level selection, automation philosophy, and verification/retrospectives. Apply optimal collaboration patterns when using any AI agent. Also the owner of the retired `agent-development-principles` name (merged 2026-09-19). Triggers on: agent principles, agentic development principles, AI collaboration principles, context-management strategy, how to work with coding agents.

Operate and extend barretlee/agent-pulse, the evidence-backed AI industry intelligence system: inspect source catalog and lifecycle, collect and normalize signals, bind evidence, cluster Events, evaluate system health, generate Scout hypotheses, export the privacy-safe public site, and verify release gates. Use when the user asks to run, configure, debug, extend, or explain Agent Pulse, its collectors, Control Room, narratives, Scout, or GitHub Pages output. Triggers on: agent-pulse, Agent Pulse, evidence-backed intelligence, source catalog, signal collection, Event clustering, source audit, Scout opportunity, public export, weekly brief, or AI industry intelligence pipeline.

Compact the current conversation into a handoff document for another agent to pick up. Use when the user says handoff, is running out of context, or wants a fresh agent to pick the work up. Triggers on: handoff, hand this off, compact this conversation.

Design domain-specific agent teams, define specialized agents, and generate the skills they use. Use when you need to decompose a complex project into coordinated multi-agent teams, choose the right architecture pattern (pipeline, fan-out/fan-in, expert pool, producer-reviewer, supervisor, hierarchical delegation), generate .claude/agents/ and .claude/skills/ files, or validate and iterate on generated harnesses. Triggers on: harness, build a harness, design agent team, agent team architecture, multi-agent skill generation, set up harness, harness engineering, domain agent team, harness for this project.

Install, configure, and operate Headroom, the local-first context optimization layer for coding agents. Use when the user needs proxy-level compression, Headroom MCP tools, persistent Claude/Codex/OpenCode routing, or a code-work policy that combines Headroom health with Graphify preflight and context-aware Ponytail minimization. Triggers on: headroom, headroom proxy, headroom deploy, headroom wrap, headroom mcp, headroom doctor, context compression, token savings, context budget, or Headroom code policy.

Operate Panniantong/Agent-Reach, the MIT-licensed installer, doctor, and routing layer for platform-specific internet research tools. Use when the user names Agent Reach, agent-reach, its channel doctor, optional backend setup, OpenCLI integration, cookie configuration, transcription, or an Agent Reach update or uninstall. Choose fit-check, offline preflight, scoped setup, channel routing, credential handoff, transcription, or maintenance. Preserve existing skills: upstream installation can replace the same-named agent-reach skill. Require explicit approval for installs, credential access, browser/session integration, Docker changes, paid uploads, and deletion. Route generic crawling to scrapling and existing authenticated browser tasks to the browser or platform skill.

Production agent SLA discipline — MAX_LOOPS, MAX_TOKENS, MAX_COST, MAX_LATENCY, MAX_TOOL_CALLS, circuit breakers, sovereignty boundaries, degradation modes. Use before deploying any autonomous loop to production.

Tool registry, capability discovery, schema-first routing, router policies and fallbacks. Use when an agent has more than ~15 tools, multiple MCP servers, or needs dynamic tool selection based on context.

Plan and improve day-to-day AI coding-agent workflow across Claude Code, Codex, Gemini CLI, and MCP-heavy repos: session startup, context recovery, fast repo loops, runtime verification, worktree use, and multi-agent handoffs. Use when the user wants a practical operating workflow, even if they ask in shorthand like shortcuts, session reset, agent setup, MCP usage, parallel agents, or better daily flow. Triggers on: agent workflow, productivity, context management, session recovery, worktree, MCP, multi-agent.

Exact rendered-UI feedback router → choose copy-paste review, watch-loop sync, self-driving critique, or platform setup. MCP: npx add-mcp "npx -y agentation-mcp server"

Discover, select, and safely install focused skills from abagames/agentic-gamedev-skills. Use when a user explicitly wants that upstream collection, needs to choose among its mini-game design, Godot, crisp-game-lib, presentation, telemetry, or agent-workflow skills, or wants a pinned selective bundle install. Triggers on: agentic-gamedev-skills, abagames skills, install gamedev skill bundle, designing-mini-games, or upstream game skill inventory.

Production engineering lifecycle for spec-first, test-driven implementation, review, and staged shipping; use for disciplined software delivery with explicit quality gates.

One-shot installer for the akillness/oh-my-gods agent skill bundle (80+ god-skills) — wraps the official install.sh with pinned URL and env knobs (PLATFORM, WITH_LANGCHAIN, INSTALL_MODE, SKIP_BACKUP) to copy the catalog into Claude Code, Codex CLI, Antigravity/Gemini, and OpenCode in one step.

Runs a scope-locked, test-backed Aider coding loop with commit hygiene; use for local-repo AI pair-programming on one small feature or bug.

Convert an article, tutorial, or prompt pack into focused reusable AgentSkills, one independent capability per skill, with portable instructions, example prompts, working demos, preview screenshots, validation, gallery updates, and a narrow commit. Use when the user asks to turn an article's prompts, tutorial sections, design patterns, interactions, or workflow ideas into complete skills rather than leaving them as prose.

Ask which skill or flow fits your situation. A router over the skills in this repo. Use when the user asks which of these engineering skills or flows fits their situation. Triggers on: ask- matt, which skill should I use, what flow fits here.

Runs directional ablation and refusal-direction analysis for open-weight models the user may modify; use to reduce benign over-refusal or measure refusal/KL trade-offs, not for training.

Drive Airship (`@airshiplabs/cli`), a CLI that puts a visual, pannable design canvas in front of a running dev server so an engineer can click an element, describe a change, and have Claude Code, Codex, or OpenCode edit the underlying source file directly — no plugin, no build config, nothing added to the project. Use when the user wants to launch a visual editor over a local dev server (`npx @airshiplabs/cli --target <port>`), pick or compare coding agents (`--agent claude|codex|opencode`) for live UI edits, sandbox agent edits with `--safe`, scaffold an `airship.config.json` (`airship init`), or diagnose a broken setup (`airship doctor`). Triggers on: "airship", "@airshiplabs/cli", "visual editor for my codebase", "click to edit UI", "design canvas over dev server", "airship doctor", "airship init", "pair Claude Code with a visual editor on localhost".

Self-hosted AI gateway with one OpenAI-compatible endpoint for multi-provider LLM, embedding, image, and audio routing, automatic fallback, load balancing, and cost optimization.

Analyze and prioritize a list of feature requests by theme, strategic alignment, impact, effort, and risk. Use when reviewing customer feature requests, triaging a backlog, or making prioritization decisions.

Drive Animato (github.com/otdnnc/Animato) as an API-key agent loop that turns a rigged .fbx/.gltf model plus a plain-text motion request into a baked animation: upload the model, build the bpy prompt, spend one LLM call with your own key, gate the generated script, run it headless, and verify the animated output. Use when the user wants text-to-animation for a 3D character, an unattended animation pipeline driven by a Gemini or OpenAI-compatible API key, or help operating a local Animato server. Triggers on: animato, text to animation, animate a rigged model, bpy animation script, blender headless keyframe, /api/chat animation, character motion from a prompt, GEMINI_API_KEY animation.

Generate an Ansoff Matrix analysis mapping growth strategies across market penetration, market development, product development, and diversification. Use when considering growth options, planning market expansion, or evaluating strategic growth paths.

Design or refactor API contracts for REST and GraphQL systems.

Use this skill when the main job is publishing or refreshing developer-facing API docs that help integrators reach first success, understand reference truth, and stay unblocked as the API evolves.

Build fully customizable, agent-ready design systems with Astryx — Meta's production design system now open source. Ships 150+ React components built on StyleX with zero styling lock-in, component swizzling, brand theming, dark mode, and CLI tooling. Use when building component libraries, design systems, UI applications, design tokens, or when teams need consistent accessible components that AI agents can understand and extend. Triggers on: astryx, design system, component library, design tokens, react components, accessible components, astryx design, stylesheets, theme customization, component composition.

Audit a website or digital experience against its supplied source references for originality and plagiarism risk. Use when Codex must compare current or historical site output with reference pages, capture packs, screenshots, copy, brands, numbers, images, assets, videos, layouts, motion, or code; raise evidence-backed red flags; distinguish common visual grammar from distinctive copying; and propose concrete fixes without making unsupported legal claims.

Audit work, verify claims with concrete evidence, and explain the result in simple grade-5 language. Use when the user asks to review, audit, check, verify, explain a change, explain a fix, summarize test results, validate whether something works, or translate technical findings into plain language for non-technical readers.

Use when you need high-quality stock-style images from Aura Assets (aura.build/assets) similar to Unsplash for design mockups and marketing: backgrounds, abstract wallpapers, architecture, portraits, and headshots. Includes a workflow for searching by tag on aura.build/assets and returns 5 real image URLs per category plus practical guidance for using different resolutions and aspect ratios.

Design or refactor product authentication setup for web apps and APIs. Use when choosing hosted/framework-native/platform-native/enterprise-add-on/self-hosted auth, sessions vs JWTs, OAuth/social login, passkeys, org/member models, callback/cookie environment setup, or SSO/SCIM rollout boundaries.

Run Karpathy-style autonomous ML search on a real training repo: choose the right mode (setup, program.md, bounded loop, results interpretation, or constrained-hardware adaptation), preserve the immutable prepare.py / 300-second / val_bpb contract, and route prompt/skill eval work away to LangSmith, Promptfoo, Braintrust, or skill-autoresearch.

AI-era curl for agent web fetching, structure discovery, and deterministic structured extraction with token-budgeted output.

Turn backend test ambiguity into one practical backend test packet. Use when the user needs API/service/repository/auth-flow coverage design, fixture or seed/reset strategy, container-vs-mock dependency choices, contract/API compatibility checks, or flaky backend-suite stabilization across local and CI.

Identify the first beachhead market segment for a product launch. Evaluates segments against burning pain, willingness to pay, winnable market share, and referral potential. Use when choosing a first market, targeting an initial customer segment, or planning market entry strategy.

Turn a completed daily UI inspiration capture into exactly five original landing-page builds, one per separate Codex task, using Sites. Use when the user asks to turn the daily inspiration references, a five-item UI prompt pack, or a dated `*-ui-inspiration-capture` article into distinct HTML landing pages while changing the source brands, names, copy, people, numbers, pricing, claims, and imagery.

Generate a Business Model Canvas with all 9 building blocks. Use when creating a business model, documenting how a business creates value, or analyzing an existing business model.

Enables persistent ultra-compressed technical communication; use for explicit brevity or token-reduction requests, except where fragments risk safety or clarity.

Operate ahujasid/blender-mcp, the MIT-licensed MCP server plus Blender addon that lets an AI client drive Blender for scene inspection, modeling, materials, rendering, asset import, and GLB or FBX export. Route one request to one mode: fit-check the project; preflight the host, addon, and port; install and wire a client such as Claude Desktop, Cursor, VS Code, Codex, or OpenCode; harden the arbitrary-code path with safe mode; control telemetry and privacy before touching client work; source or generate 3D assets under their licences; export to disk; or troubleshoot a broken connection. Use when a user wants an AI agent to control Blender. Triggers on: blender-mcp, MCP for Blender, control Blender with Claude, execute_blender_code, BLENDER_MCP_SAFE_MODE, Poly Haven, Hyper3D Rodin, Hunyuan3D, Sketchfab import, Poly Pizza, Blender addon port 9876.

Packet-first BMAD/BMM front door for idea notes, product briefs, PRDs, architecture drafts, review feedback, existing repo state, and milestone pressure. Use when the user wants to know what BMAD phase or artifact comes next, or needs a portable BMAD entrypoint before routing review, execution slicing, runtime setup, or game-production work outward.

AI-driven Game Development Studio (BMAD-GDS). Routes game projects through Pre-production, Design, Architecture, Production, and Game Testing phases using 6 specialized agents. Supports Unity, Unreal Engine, Godot, and custom engines.

Turn raw ideas into one clear pre-planning concept artifact by choosing the right framing mode: problem framing, audience and value framing, concept shaping, game concept framing, or story packaging. Use when the user needs the next artifact before PRDs, sprint plans, launch execution, or game-production routing.

Design experiments to test assumptions for an existing product — prototypes, A/B tests, spikes, and other low-effort validation methods. Use when validating assumptions, testing feature ideas cheaply, or planning product experiments.

Design lean startup experiments (pretotypes) for a new product. Creates XYZ hypotheses and suggests low-effort validation methods like landing pages, explainer videos, and pre-orders. Use when validating a new product idea, creating pretotypes, or testing market demand.

Brainstorm product ideas for an existing product using multi-perspective ideation from PM, Designer, and Engineer viewpoints. Use when generating new feature ideas, brainstorming solutions for an identified opportunity, or ideating with a product trio.

Brainstorm feature ideas for a new product in initial discovery from PM, Designer, and Engineer perspectives. Use when starting product discovery for a new product, exploring features for a startup idea, or doing initial ideation.

Brainstorm team-level OKRs aligned with company objectives — qualitative objectives with measurable key results. Use when setting quarterly OKRs, aligning team goals with company strategy, drafting objectives, or learning how to write effective OKRs.

Automates and verifies multi-step browser tasks through a clean Chrome CDP session; use when repeatable autonomous browser control is needed, not an already-open authenticated profile.

Create polished 60 fps 4:3 4K browser screen-recording style videos from Codex in-app browser captures, with browser-only crop, natural macOS cursor styling, deliberate click choreography, zoom-follow framing, ffprobe/thumbnail verification, and optional native recording compatibility checks. Use when the user asks to record or re-record browser actions, show cursor clicks and zooms, make Dribbble/UI inspiration or product demo recordings, or asks whether Codex, Playwright, or an MCP can produce a natural browser demo video.

Write and maintain release-history artifacts for shipped changes: `CHANGELOG.md` updates, release notes, migration/deprecation updates, and lightweight game patch notes. Use when the main job is turning shipped evidence into the smallest truthful release-writing packet for developers, customers, internal stakeholders, or players. Also owns the retired `release-notes` name: turn tickets, PRDs, or Git logs into user-facing notes that lead with the benefit. Triggers on: changelog, release notes, patch notes, migration update, deprecation notice, version notes, what shipped, what changed, what's new, announce product updates. Route internal specs/runbooks to `technical-writing`, API portals to `api-documentation`, end-user tutorials to `technical-writing`, rollout execution to `deployment-automation`, and launch messaging to `marketing-automation`.

Scaffold and extend chatbot-template, shadcn/ui's minimal Next.js starter for building an AI chatbot with the AI SDK, shadcn/react, shadcn/typeset, and the Vercel AI Gateway. It ships streaming markdown chat, tool calling (a server-executed GitHub repo lookup, provider-native web search, and a human-in-the-loop `ask_user` questionnaire), and one-click Vercel deploy with OIDC gateway auth. Use when the user wants to spin up a Next.js AI chat app fast, add a custom tool with a typed UI part, swap or restrict the model list, or harden the public `/api/chat` route before shipping. Triggers on: "chatbot-template", "shadcn chatbot", "AI SDK chat app", "Next.js AI SDK starter", "Vercel AI Gateway chat", "streaming chat with shadcn/typeset", "ask_user questionnaire tool", "shadcn/react message scroller".

Claude Code hook library — pre-built PreToolUse/PostToolUse hooks for common guardrails, auto-formatting, workflow automation

Make any software agent-native with HKUDS CLI-Anything — four routed modes: install ready-made harnesses via CLI-Hub (cli-hub list/search/info/install/launch), give agents the cli-hub-meta-skill for autonomous discovery, generate a new harness from any codebase/repo via the 7-phase /cli-anything pipeline, or iterate with :refine/:test/:validate; 40+ harnesses, 2,461 tests, Click CLIs with REPL + --json.

Turn a cleanup packet into one behavior-preserving refactor brief. Use when simplifying a messy function/component/service, freezing behavior before touching fragile legacy code, splitting a cleanup-heavy diff into reviewable slices, or planning a repeated migration / codemod without changing intended behavior. Route diagnosis to debugging, review judgment to code-review, validation-program design to testing-strategies, bottleneck-led tuning to performance-optimization, and pure symbol inventory to codebase-search.

Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use when the user wants to review a branch, a PR, work-in- progress changes, or asks to "review since X".

Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.

Route repo-navigation requests to one search packet before editing: exact-text, symbol/indexed, structural, config/content, hosted search, or graph/path trace. Use when the user asks where something is defined or referenced, which files own config/content surfaces, or what must be inspected before a change.

Routing front door for a structured, human-in-the-loop deep-research workflow (Weizhena/Deep-Research-skills) — turn a topic into an extensible outline, fan out parallel web-search agents to investigate each item into validated JSON, then render a complete markdown report. One skill, 4 reference pipelines: outline (research / add-items / add-fields), deep (parallel per-item investigation + field-coverage validation), report (TOC + per-field markdown), web-search (research agent + 5 routed source modules). Plugin: npx skills add https://github.com/akillness/jeo-skills --skill deep-research

Drive CodeBurn, a free open-source local-first CLI/TUI/web/menubar tool that reads the session files already on disk from 40 AI coding tools (Claude Code, Codex, Cursor, Gemini CLI, Grok, OpenCode, and more) and breaks down token usage and dollar cost by task, model, tool, and project. Use when the user wants to see where their AI coding spend went, find and fix token waste in a Claude Code / agent setup, cap a session's budget before it runs away, compare which model is actually worth its price, check whether AI spend shipped or was reverted, or wire live usage/savings data into an agent over MCP. Triggers on: "codeburn", "npx codeburn", "AI token usage", "AI coding cost", "where did my Claude spend go", "codeburn optimize", "codeburn guard", "codeburn compare models", "codeburn yield", "token waste in CLAUDE.md", "AI spend dashboard".

Turn any GitHub repo, local folder, PR, or markdown/Obsidian vault into an interactive architecture map with CodeFlow — a zero-build single index.html browser app (React 18 + D3.js from pinned CDNs) that runs 100% client-side with no backend and no data collection. Pick an input (public repo, private repo with a local token, local files, PR URL, markdown vault), choose a visualization mode (folder/layer/churn/blast), and read the analysis — dependency graph, blast radius, code ownership, heuristic security scanner, pattern/anti-pattern detection, A–F health score, activity heatmap, PR impact — then export JSON/Markdown/text/SVG/PDF or wire the self-updating CodeFlow Card SVG onto a README.

Perform cohort analysis on user engagement data — retention curves, feature adoption trends, and segment-level insights. Use when analyzing user retention by cohort, studying feature adoption over time, investigating churn patterns, or identifying engagement trends.

Assist with Colibri: pure-C LLM inference engine for running GLM-5.2 (744B MoE) on consumer machines with ~25 GB RAM. Use when setting up, building, converting models, running inference, configuring expert streaming and caching, optimizing speculative decoding (MTP), GPU integration, and integrating Colibri into production pipelines. Includes build setup, model download & conversion, chat/inference modes, performance tuning, and API integration patterns.

Create sales-ready competitive battlecards comparing your product against a specific competitor — positioning, feature comparison, objection handling, and win/loss patterns. Use when preparing sales teams, creating competitive materials, or responding to 'why not competitor X?'

Analyze competitors with strengths, weaknesses, and differentiation opportunities. Identifies direct competitors and maps the competitive landscape. Use when doing competitive research, preparing a competitive brief, or finding differentiation opportunities.

Free offline desktop video/image compression (Tauri+React) — batch compress, trim/split, convert formats, embed subtitles. Install: brew install --cask codeforreal1/tap/compresso

Create a Product Requirements Document using a comprehensive 8-section template covering problem, objectives, segments, value propositions, solution, and release planning. Use when writing a PRD, documenting product requirements, preparing a feature spec, or reviewing an existing PRD.

Create or refresh hierarchical AGENTS.md documentation for Claude Code, Codex/OMX, Gemini, and Antigravity/OMA projects, preserving manual notes while excluding runtime state such as root .omc, .omx, .survey, .codex, and generated build folders.

Plan and review release execution for web, backend, and fullstack systems: preview releases, staging-to-production promotion, rollout strategy, post-deploy verification, rollback response, and release-hardening checklists. Use when the system can already build and the main job is shipping or recovering a release safely, choosing between preview / promotion / canary / rollback paths, or tightening deploy gates around health checks and sign-off. CI workflow and release-job authoring belongs here too. Route machine/runtime setup to `system-environment-setup`, long-lived telemetry design to `monitoring-observability`, and Vercel-specific operations to `vercel-deploy`.

Gmail customer support triage with mandatory full-thread read-back before every reply or follow-up, a draft-only default, per-draft Codex project threads, approval-gated final closure replies with eligible optional Trustpilot review invitations, and approval-gated operations for verified DreamCut Discord invite requests. Use when the user asks to run the customer email automation, check unread/recent support emails, review a customer follow-up, prepare or send Gmail replies, close a positively confirmed resolved case, continue a support handoff, create agent/project threads for drafted follow-up, or fulfill DreamCut Discord access requests.

Create an end-to-end customer journey map with stages, touchpoints, emotions, pain points, and opportunities. Use when mapping the customer experience, identifying friction points, improving onboarding, or visualizing the user journey.

Verify customer support work against the selected customer-case flow, full-thread evidence, account and canonical-thread matching, authority, financial approval, customer communication, eligible optional Trustpilot review-request rules, post-action read-back, send confirmation, thread-wide archive, and closure requirements. Use after every SaaS support triage, account/access or unsupported-platform response, duplicate or resolved acknowledgement, positive-confirmation closure reply, cancellation, failed-payment cancellation, refund, accidental-renewal case, support handoff, draft, send, or archive task before reporting completion. Triggers on: verify support case, support closure check, cancellation verification, refund verification, accidental-renewal case, duplicate/resolved acknowledgement, account-access ticket, Trustpilot eligibility, thread read-back before marking done.

Create a recurring daily UI inspiration capture. Use when the user asks to run, refresh, package, or validate dated UI inspiration bundles, especially for `articles/YYYY-MM-DD-ui-inspiration-capture/` outputs, Framer/Dribbble landing-page inspiration, motion-study screenshots/videos, AI-builder prompts, duplicate checking, or converting a project runbook into repeatable workflow.

Install, operate, and troubleshoot 0ceal0t/Dalamud-VFXEditor for user-owned Final Fantasy XIV visual-effect, animation, sound, physics, UI, texture, material, model, and shader files. Use when a user asks about `/vfxedit`, AVFX, PAP, TMB, SCD, Dalamud plugin setup, effect replacement, export handoff, beta repositories, or VFXEditor source builds. Triggers on: Dalamud VFXEditor, /vfxedit, .avfx, .pap, .tmb, .scd, FFXIV VFX mod, or VFXEditor beta.

Analyze datasets to extract insights, identify patterns, and generate reports.

Design storage-model and migration-safety packets for relational, document-heavy, and hybrid data systems. Use when the user needs entity ownership, constraints, indexes, staged schema changes, or queryable-vs-flexible field decisions across backend/fullstack products, internal ops tools, marketing/customer-data workflows, or game/live-ops systems. Route API contracts to api-design, auth-owned identity/session modeling to authentication-setup, verification to backend-testing, and reporting/telemetry follow-through to looker-studio-bigquery or monitoring-observability.

Use when you need design-first, spec-driven, skimmable prompts for UI generation. Covers prompt structure, constraints, variations, typography/spacing rules, and iteration workflow for consistent UI outputs.

Define or refactor a shared frontend UI system before polishing one screen at a time. Use when the user needs token governance, visual-language rules, primitive naming, page-system direction, cross-product consistency, or a design handoff that should stay coherent across landing pages, dashboards, forms, and component libraries, including the reusable primitive / slot / variant APIs that system owns. Not for responsive layout-only fixes, accessibility-only remediation, or broad UI critique. Triggers on: design system, design tokens, visual language, component library foundations, UI governance, cross-product consistency, landing page plus dashboard system, primitive naming, token policy, component API architecture.

Apply anti-generic frontend design rules bundled from Leonxlnx/taste-skill, the MIT-licensed anti-slop ruleset for landing pages, portfolios, editorial pages, and redesigns. Use when output looks templated, AI-default, or "sloppy" and the user wants a deliberate design read, explicit DESIGN_VARIANCE / MOTION_INTENSITY / VISUAL_DENSITY dials, a real design system choice, an AI-tells sweep including the em-dash ban, a redesign audit that preserves brand equity, or the pre-flight check before shipping. Even if the user does not say taste-skill, also triggers on: looks like AI made it, generic landing page, purple gradient hero, make it less templated, Awwwards feel, Linear-style, redesign without losing the brand. Not for dashboards, data tables, wizards, or admin UI. Route shared token governance to `design-system`, layout adaptation to `responsive-design`, accessibility remediation to `web-accessibility`, and visual-recipe lookup to `web-design`.

[devup-ui] Routing-first front door for Devup UI — a zero-runtime CSS-in-JS library whose Rust + WebAssembly preprocessor extracts every style at build time, so no styling JavaScript ships at runtime. Use to adopt/integrate Devup UI, wire the build-time plugin into Next.js/Vite/Rsbuild/Webpack/Bun, write styles with Box/css props or the styled-components-compatible styled() API, set up type-safe theming with devup.json, or migrate off styled-components/Emotion/Tailwind/Panda/vanilla-extract to zero runtime.

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a CONTEXT.md, or recording or editing an ADR.

Draft a detailed Non-Disclosure Agreement between two parties covering information types, jurisdiction, and clauses needing legal review. Use when creating confidentiality agreements or preparing an NDA for a partnership.

Install, route, and operate zenstory-ai/drama-skills, the MIT-licensed 10-skill creator-first suite for Chinese short dramas and motion comics. Use when the user wants to import or troubleshoot the suite; initialize or resume a filesystem project; analyze a novel; develop an adaptation; write episodes; build visual assets; produce image prompts, storyboards, or video prompts; review a project; open its local Dashboard; or run confirm-gated image, video, TTS, or music production. Route each request to the correct `short-drama-*` owner while preserving the five-document episode contract. Triggers on: drama-skills, zenstory-ai/drama-skills, short-drama, Chinese short drama, motion comic, creator-first drama workflow, 剧本, 视觉设定, 分镜, 图片提示词, 视频提示词. Route generic programmable-video work to `video-production`, webtoon panel production to `webtoon-harness`, and the OpenStory codebase to `openstory`.

Design, implement, budget, and verify readable real-time game visual effects: particles, smoke/fog, lightning, fire, trails, impacts, bloom/glow, and layered spell timelines. Use when a user needs a VFX spec, lifecycle, engine handoff, particle pooling, effect composition, reduce-motion fallback, or frame-budget diagnosis. Triggers on: game VFX, spell effects, particle system, impact burst, procedural lightning, smoke, bloom, magic aura, or VFX performance.

Turn a 5-second voiceover script, opinion statement, or abstract concept into an editorial halftone paper-collage assemble-from-empty B-roll video. Use when the user requests "collage b-roll", "paper collage B-roll", "halftone collage", "collage video", "gbro-collage-broll", or asks to convert voiceover lines into animated editorial collage visual metaphors. Enforces a 3-gate approval workflow (metaphor confirmation -> still frame confirmation -> video generation via Gemini Omni Flash).

Turn natural-language descriptions into editable `.drawio` diagrams and export them to PNG / SVG / PDF / JPG via the native draw.io desktop CLI, or turn an existing codebase (Python / JS-TS / Go / Rust) into an auto-laid-out structure diagram. Wraps Agents365-ai/drawio-skill: 6 diagram presets (ERD, UML class, sequence, architecture, ML/DL, flowchart), search across 10,000+ official AWS/Azure/GCP/Cisco/K8s/UML/ BPMN shapes, 321 AI/LLM brand logos, vision self-check + auto-fix, and a 5-round iterative refinement loop. No MCP server, no daemon — runs from a single SKILL.md and the draw.io CLI. Use when the user wants polished, precise, exportable diagrams or wants to visualize code structure. Triggers on: drawio, draw.io, drawio diagram, architecture diagram, ERD, UML diagram, sequence diagram, flowchart, network diagram, visualize codebase, code structure diagram, class hierarchy, export diagram png/svg/pdf, AWS/Azure/GCP icon, draw.io shapes.

Generate realistic dummy datasets for testing with customizable columns, constraints, and output formats (CSV, JSON, SQL, Python script). Use when creating test data, building mock datasets, or generating sample data for development and demos.

Install, configure, verify, update, or recover affaan-m/ECC, the MIT-licensed agent harness with native Claude Code and Codex plugins plus selective adapters for other supported clients. Use when the user names ECC, Everything Claude Code, ecc-universal, ecc@ecc, `ecc setup`, or an ECC plugin, hook, rule, skill, or harness installation. Route generic agent-team design to `harness`, delivery workflow selection to `agentic-skills`, and BMAD phase selection to `bmad`.

Generate ElevenLabs text-to-speech audio from scripts or inline text using local voice profiles. Use when the user asks for ElevenLabs, text-to-speech, TTS, narration, voiceover, speech audio, or voice generation; load voice names, voice ids, emails, owners, and account-specific defaults only from local config outside the skill.

Explain a topic, codebase, document, concept, or error at a specific audience's knowledge level and decision context. Use for ELI5, explain like I'm five, explain this to my manager, parent, child, partner, or team, break it down, dumb it down, make it understandable, or simplify it for a named age, grade, education level, job role, or relationship. This skill owns pure audience-adaptive explanation. Route requests that first require auditing or verifying a claim, change, fix, or test result to `audit-verify-explain-grade-5`; route tutorials, runbooks, and help-center content to `technical-writing`.

Organize application environment configuration: `.env` files, env precedence, typed env validation, secret handoff, framework-specific env rules, and config drift between local, staging, CI, and production. Use when the user needs help structuring environment variables, validating required config, separating public/private env values, or cleaning up env-file sprawl. This is the narrower app-config compatibility skill. Route broader runnable-machine, Docker, devcontainer, onboarding, and local-service setup work to `system-environment-setup`.

Routing-first Fabric operator skill for reusable named AI transforms over stdin, files, transcripts, notes, logs, and cleaned web text. Use when the user wants Fabric patterns, custom pattern packs, shell-pipe composition, or `fabric --serve` workflows — not generic one-off chat prompting, repo-aware coding, or fully deterministic automation.

Design or refactor project structure around the right boundary unit: feature, shared layer, route segment, or workspace package. Use when a repo feels scattered, a team needs naming/import conventions, or the user must choose between type-based, feature-based, framework-colocated, and workspace layouts. Triggers on: file organization, folder structure, project structure, reorganize repo, feature folders, shared vs feature code, where should this file live, apps/packages split, and project layout refactor.

Implement a piece of work based on a spec or set of tickets. Use when an agreed spec or ticket set should now be built. Triggers on: implement, build this spec, work these tickets.

Implement, tune, debug, or validate soft wall-aware fog of war and gameplay perception in Three.js action games. Use for orthographic or isometric visibility masks, obstacle-aware line of sight, player and enemy vision ranges, hidden-enemy targeting rules, fog shader artifacts such as spokes or seams, mobile ray budgets, lifecycle and menu-state integration, and deterministic fog-of-war tests.

Discover and install Agent Skills from the public skills.sh ecosystem using the `npx skills` CLI. Route one request to search, quality triage, install, update, or authoring handoff. Use when the user asks "is there a skill for X", "find me a skill", "how do I do X" where a published skill may already exist, wants to browse the skills.sh leaderboard, or wants to install from `vercel-labs/agent-skills`, `anthropics/skills`, or another GitHub owner. Verify install count, source reputation, and repository stars before recommending, and require confirmation before any install, global `-g` write, or `--yes` run. Route browsing the local jeo-skills catalog to `jeo-skill`, in-repo skill retrieval and ranking to `openspace`, authoring a new skill to `write-a-skill`, and spec compliance to `skill-standardization`.

Operate Firebase from the terminal with `firebase-tools`: install/auth the CLI, bootstrap `firebase.json` / `.firebaserc`, run the Emulator Suite, deploy Hosting / Functions / rules / App Hosting, manage preview channels, and handle Firebase admin tasks like auth import/export, Remote Config, App Distribution, and Extensions. Use when the job is Firebase platform/project operation through the CLI. Triggers on: firebase deploy, firebase init, firebase emulators, firebase hosting, firebase functions, firebase firestore, firebase database, firebase auth import, firebase remote config, firebase app distribution, firebase extensions, firebase apphosting, firebase dataconnect, firebase cli, firebase-tools, deploy firebase, firebase preview channel, firebase login, firebase use, firebase target apply. Route backend AI workflow orchestration to `genkit` and direct in-app SDK integration to `genkit` (`client-ai-logic` mode).

Triage Unity and Unreal Engine editor, build, package, cook, compile, and CI logs into the first actionable failure, likely subsystem, and next debugging steps. Use when a game project fails to build, cook, package, import assets, compile scripts/code, or throws noisy editor/runtime errors, even if the user only shares raw log text or says "Unity build failed", "Unreal cook error", "Editor.log", "UHT/UBT failed", or "packaging started breaking after asset moves".

Turn messy Unity or Unreal build/release automation into one bounded game-pipeline packet after naming the signal tier first: fast branch-gate CI, nightly/package- candidate builds, or release/certification candidates. Use when a game team needs to design or repair GitHub Actions, Unity Build Automation, Jenkins, TeamCity, or similar CI for engine builds — especially when they mention flaky packaging, cache superstition, giant build-job blobs, slow cook / package cycles, artifact confusion, SDK/signing drift, or manual candidate promotion that should become reproducible. Route one red log first-pass diagnosis to `game-build-log-triage`.

Triage mixed game demo and playtest feedback into a prioritized fix brief, weighted evidence summary, and next artifact recommendation. Use when a team has playtest notes, Steam Playtest responses, creator or streamer demo reactions, survey comments, wishlist/context signals, bug lists, or performance findings and needs to decide what to fix first before the next build, festival, or launch beat, even if they only say "sort our playtest feedback", "what should we fix before Next Fest", "players are confused", "streamers bounced off the demo", or "turn these demo notes into priorities".

Generate multiple original brand campaign worlds from a supplied visual reference while controlling how close the new work feels without copying protected signature elements. Use when a user provides a brand identity image, poster, editorial campaign, moodboard, or generated concept and asks for inspired alternatives, several distinct brands in one visual family, a closer-to-reference V2, exact wordmarks inside images, or originality-safe image-generation prompts and outputs.

Route Firebase AI feature work into either direct app/client Firebase AI Logic SDK integration or a server-owned Genkit workflow. Use when a web, mobile, backend, or full-stack feature needs model calls, typed outputs, reusable flows, tools, retrieval, prompt files, evals, observability, or deployment. Choose client-ai-logic, flow-foundation, tool-and-agent, retrieval-and-prompt, evaluation-and-observability, deployment-runtime, or comparison-or-fallback; route Firebase platform/operator work to `firebase-cli` and broad framework comparisons to `survey`.

Search and download specific files/folders from GitHub repositories directly from terminal using ghgrab, without full clone. Covers install, interactive browsing, release asset download, and automation-safe usage patterns.

Turn a game-design question into one falsifiable theory packet instead of a framework collage. Use when a designer must explain why a mechanic may create a player experience, analyze mechanics-dynamics-aesthetics, test a core loop, inspect choice, uncertainty, progression, resource flow, mastery, autonomy, competence, or relatedness, compare design variants, or turn conflicting playtest interpretations into a bounded prototype. Choose one primary lens, state its limits, separate observation from assumption, map the causal chain, define counterevidence, and validate `game-design-hypothesis.json`. Triggers on: game design theory, MDA, mechanics dynamics aesthetics, player motivation, core loop theory, dominant strategy, reward loop, flow, Bartle types, fun analysis, design hypothesis, or why this mechanic works.

Diagnose and tune the response chain from player intent through input, simulation, visible motion, camera, animation, VFX, audio, haptics, UI, and recovery for an existing game mechanic. Use when movement, combat, jumping, firing, hits, pickups, menus, or abilities work mechanically but feel delayed, weak, weightless, noisy, inconsistent, nauseating, or unresponsive; when the user asks for game feel, juice, punch, hit stop, freeze frames, screenshake, squash and stretch, impact feedback, input latency, coyote time, or feedback layering; or when a team needs a baseline-versus-variant capture and an accessibility-safe tuning contract. Measure before tuning, change one causal variable, preserve simulation truth, and validate `game-feel-contract.json`.

Route Unity and Unreal frame-time complaints into one bottleneck-first profiling brief. Use when the main job is interpreting profiler screenshots, `stat unit` / `stat gpu` output, benchmark-route complaints, or Steam Deck / target-device review packets; choosing the smallest useful next capture; naming one primary bottleneck family; and deciding whether to stay with quick packets, move to an engine-native profiler, or escalate further. Route generic app/service tuning to `performance-optimization`, build/editor/package failures to `game-build-log-triage`, broader game-production coordination to `bmad-gds`, and mixed demo/community feedback to `game-demo-feedback-triage`.

Install, configure, and troubleshoot Citedy's game-sounds feedback audio for Claude Code and supported CLI environments. Use when a user wants coding-agent event sounds, sound-pack rotation, volume or event toggles, playback checks, custom packs, or the game-sounds CLI/plugin. Triggers on: game-sounds, coding sounds, Claude hook sounds, task-complete sound, sound pack rotation, or @citedy/game-sounds.

Five-role game production studio harness: director, numeric-balance designer, revenue-band PM, verification-strict programmer, archetype-rotation QA. Runs the 3-stage operating cycle (concept/presentation/core build → balance/core-loop/novelty → ops stability/play impact) behind 8 numeric quality gates with survey-grounded trends, signed designer↔PM negotiation records, and QA broadcast discipline. Writes the repository rule file (CLAUDE.md/AGENTS.md) so the contract outlives the session, and keeps one live `_workspace/current/` beside a read-only archive.

Design and review an engine-neutral game UI/UX contract for HUDs, menus, inventories, shops, maps, settings, overlays, tutorials, notifications, and controller flows. Use when a game interface must define player decisions, information hierarchy, persistent versus contextual HUD state, screen-stack behavior, keyboard/gamepad/touch input, focus order, back behavior, safe areas, responsive layout, scalable text, localization, accessibility, event-driven data binding, or cross-device verification; when the user asks for game UI design, game UI UX, HUD usability, controller navigation, or Three.js game UI planning; or when a mockup must be reconciled with a real runtime. Produce and validate `game-ui-contract.json`, then route visual concepts and engine widgets to their specialist skills.

Decide when Git submodules are the right external-repo boundary, then choose one safe operator flow: add and pin, bootstrap recursively, sync to the recorded commit, advance a tracked branch, edit inside the submodule without detached-HEAD surprises, remove cleanly, or configure CI/hosted-platform checkout constraints. Use when the user asks about `.gitmodules`, `git submodule`, recursive clone/setup, pointer updates, detached HEAD, private submodules in CI, GitHub Pages submodule limits, or submodule vs subtree/vendoring/package delivery. Not for generic Git history cleanup or package-manager dependency delivery.

Route local Git work into the safest next move: branch hygiene, selective staging, commit cleanup, merge-vs-rebase choice, conflict resolution, lease-safe pushes, and recovery from resets or bad history edits. Use when the user needs help preparing a branch, cleaning up commits, syncing with an updated base, resolving local Git conflicts, pushing rewritten history safely, recovering lost commits, or getting a diff ready for review. Not for hosted PR review, repo administration, or sprint planning.

Convert noisy GitHub repository search results into recommendation-grade candidate lists with explicit metadata, freshness, traction signal, provenance labels, and rollback-safe reporting for maintenance PR workflows.

Route goalflow (wanmol/goal-flow) work — a LangGraph framework that combines workflow graphs with agent loops — into exactly one mode: fit check, transpiling a Dify DSL export into runnable LangGraph Python, authoring workflow nodes and edges, building an `agent_kit` loop with middleware and a harness, wiring the serving layer (data adapters, SSE streaming, HITL, Redis/MySQL, API-key registration), or running the pre-publish security gate. Use when the user wants Dify's visual design without Dify's runtime, a graph node that hosts an agent loop, an OpenAI-compatible wire protocol over their own workflows, or prompt-injected `SKILL.md` capabilities. Triggers on: goalflow, goal-flow, dify to langgraph, dify transpiler, dify DSL export, BaseWorkflow, agent_kit, AgentBaseNode, DataAdapter, chunk processor, HITL interrupt, dify2langgraph. Route plain graph-API questions to `langgraph-fundamentals`.

Generate images using Codex's ChatGPT backend with zero production dependencies. Reuses existing local Codex authentication (~/.codex/auth.json) — no new credentials needed. Supports CLI (gti command), Node.js library, and Python SDK. Accepts text prompts with optional reference images (PNG/JPG/GIF/WebP). Includes dry-run mode and debug output. Triggers on: god-tibo-imagen, gti, image generation, codex image, chatgpt image, ai image, gpt image generation.

Drive Godogen (htdt/godogen), the MIT-licensed publish-time generator that turns a game description into an autonomous Claude Code or Codex build for Godot 4 C#, Bevy Rust, or Babylon.js TypeScript. Route one request to one mode: preflight the toolchain and API keys; publish a fresh game repository or safely refresh a matching existing runtime with `./publish.sh --engine ...`; run the build and prove it from the live game or a 15-20s recording; budget paid Gemini, Grok, and Tripo3D asset generation; apply engine-specific build and capture rules; troubleshoot rendering and capture failures; or contribute through the issue-first upstream process. Use when the user wants an agent to build a playable game end to end with Godogen. Triggers on: godogen, htdt/godogen, publish.sh --engine, autonomous game development, Godot C# agent build, Bevy agent build, Babylon.js agent game, asset-gen, Tripo3D rig, proof video.

Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick. Use when the user wants architecture improvement candidates surfaced before choosing one. Triggers on: improve the architecture, find deepening opportunities, where should we refactor.

Create a structured customer interview script with JTBD probing questions, warm-up, core exploration, and wrap-up sections. Follows The Mom Test principles — no leading questions, no pitching, focus on past behavior. Use when preparing for user interviews, creating interview guides, or planning discovery research.

Guide performance profiling for Apple platform apps with Instruments, Xcode diagnostics, and MetricKit. Use when investigating app hangs, stutters, high CPU, memory leaks, memory growth, OOM crashes, slow launch, battery drain, thermal issues, App Store performance readiness, or when adding os_signpost and measurement hooks.

Plan and execute Google Workspace operations across Docs, Sheets, Slides, Drive, Gmail, Calendar, Forms, Chat, and Admin SDK by choosing the right surface first: Apps Script, direct REST API, or admin-only APIs. Use when the user needs to automate a Workspace workflow, edit or create Workspace files, send Gmail, schedule calendar events, process Forms/Sheets pipelines, manage Drive sharing, or handle Workspace admin tasks; even if they only mention a spreadsheet, document, shared drive, inbox workflow, approval flow, or domain user change. Triggers on: Google Doc, Google Sheet, spreadsheet, Slides, Drive folder, Gmail, calendar event, Forms, Apps Script, Admin SDK, shared drive, Google Chat, workspace automation, domain user, send email, approval flow.

Identify grammar, logical, and flow errors in text and suggest targeted fixes without rewriting the entire text. Use when proofreading content, checking writing quality, or reviewing a draft.

Drive Graphify from its CLI to build, refresh, query, export, and serve a durable code/corpus knowledge graph. Use when the user wants `.graphify/GRAPH_REPORT.md`, `graph.json`, `graph.html`, `graphify update`/`summary`/`query`/`path`/`explain`/`tree`, change-aware review context, git-hook or watch-based refresh, a stdio MCP graph server, or an install into jeo, jeopi, gjc, opencode, Claude, Codex, or Gemini. Also covers the honest structural fallback when native extraction is empty or misleading. Route simple locate/reference work to `codebase-search`, narrative knowledge-base work to `llm-wiki`, and project-memory handoff to `opencontext`. Triggers on: graphify, graphify update, graphify query, knowledge graph CLI, GRAPH_REPORT.md, graph.json, codebase graph, graph refresh, graphify install, graphify serve, review context, affected flows.

A relentless interview to sharpen a plan or design. Use when the user wants a plan or design stress-tested. Triggers on: grill me, grill this, stress-test my plan.

A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go. Use when the grilling session should also leave ADRs and glossary entries behind. Triggers on: grill with docs, grill and document, grill this and write the ADR.

Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress- test their thinking, or uses any 'grill' trigger phrases.

Identify growth loops (flywheels) for sustainable traction. Evaluates 5 loop types: Viral, Usage, Collaboration, User-Generated, and Referral. Use when designing growth mechanisms, building product-led traction, or understanding how growth loops work.

Identify the best GTM motions and tools across 7 motion types: Inbound, Outbound, Paid Digital, Community, Partners, ABM, and PLG. Use when selecting marketing channels, choosing between inbound and outbound strategy, or planning cross-channel campaigns.

Create a go-to-market strategy covering marketing channels, messaging, success metrics, and launch timeline. Use when planning a product launch, creating a GTM plan from scratch, or defining a launch strategy for a new market.

Create and extend complete visual brand systems through the Higgsfield CLI and bundled deterministic local tooling: palettes, SVG logo marks, typography, mockups, social graphics, packaging, signage, merchandise, posters, presentation decks, and editable PPTX/PDF brandbooks. Preserves official supplied assets, persists approvals locally, and regenerates only dependent outputs. Use when: "create a brand kit", "make a visual identity", "design a logo and brandbook", "apply this logo to branded assets", "make packaging or signage", or "extend our existing branding". Chain with higgsfield-generate for general image production and Marketing Studio brand-kits when importing website metadata for ads. NOT for unbranded image generation (use higgsfield-generate), product catalog photography (use higgsfield-product-photoshoot), website implementation (use higgsfield-websites), or native Figma/Canva/PSD/AI delivery.

Generate images/videos/3D assets/audio via Higgsfield AI. Defaults: GPT Image 2.5 for image/design/text, Seedance 2.5 for video, Nano Banana 2/Lite/Pro for character/reference images, Marketing Studio for ads, Seed Audio 1.0 for audio. Use when: "generate an image", "make a video", "animate this photo", "image-to-video", "edit/stylize/remix this image", "reframe this video", "edit this video from a sketch", "create a 3D model/GLB", "create a sound effect", "make music", "text-to-audio", "create an ad", "make a UGC video", "unboxing", "presenter video", "import product from URL", or "analyze video virality". Supports generic generation, workflows, Marketing Studio, and Virality Predictor. Chain with higgsfield-soul-id for face/identity consistency. NOT for: Soul training, brand systems/brandbooks (use higgsfield-brandkit), photoshoots, cards, YouTube thumbnails (use higgsfield-youtube-thumbnail), explainers (use higgsfield-video-explainer), playable games/assets (use higgsfield-websites), or TTS.

Generate marketplace product image cards through Higgsfield: compliant main image, secondary product images, and A+ style content modules. Use when the user asks for marketplace listing images, product detail cards, secondary product images, product infographics, lifestyle listing shots, A+ style content, marketplace image sets, or sales-ready product visuals. Backend owns marketplace compliance references and prompt templates; this skill only routes user intent to the CLI. NOT for generic brand product photography without marketplace/listing context (use higgsfield-product-photoshoot), video generation or UGC ads (use higgsfield-generate), or Soul Character training (use higgsfield-soul-id).

Browse, group, relate, and selectively install the jeo-skills catalog through the lightweight `jeo-skill` CLI. Use when the user wants skills organized by web, infrastructure, game, creative media, CLI tools, AI/agents, engineering, research, business, or utilities; needs a frontend/backend/game-audio/game-VFX subcategory; wants overlapping skills connected instead of duplicated; or wants a category, bundle, or named skills installed without copying the full repository.

Generate brand-quality product images through Higgsfield product-photoshoot prompt enhancement on GPT Image 2 / gpt_image_2. Entry point for professional brand/product visuals. Use when: "product photo", "studio shot", "lifestyle image", "Pinterest pin", "hero/banner", "carousel", "ad creative", "Meta ads", "virtual try-on", "model wearing", "person holding product", "closeup with hands", "levitating/floating/splash product", "CGI/surreal product", "restyle", "seasonal/aesthetic variation", or any product, brand, or paid-social creative. Modes: product_shot, lifestyle_scene, closeup_product_with_person, moodboard_pin, hero_banner, social_carousel, ad_creative_pack, virtual_model_tryout, conceptual_product, restyle. Backend assembles the final prompt; never freehand it. NOT for: no-product text-to-image (use higgsfield-generate), branded avatar video (use higgsfield-generate Marketing Studio), marketplace listing cards (use higgsfield-marketplace-cards), Soul Character training (use higgsfield-soul-id).

Train a Soul Character — a personalized model on a person's face that Higgsfield uses for identity-faithful image and video generation. Use when: "create my Soul", "train my face", "make my digital twin", "build me an avatar", "learn my appearance", "create a character of me", "set up identity for video", "I want my face in generated images". Chain: train Soul (one-time, returns reference_id) → use in higgsfield-generate via `--soul-id <id>` with models like `text2image_soul_v2` or `soul_cinema_studio`. NOT for: one-shot face swaps (use higgsfield-generate with --image), named-character / non-photo avatars (use higgsfield-generate with prompt).

Build a complete non-photoreal narrated explainer or story video from ordered 10-second blocks: one narrator, one universal style key, one Seed Audio take and one Gemini Omni clip per block, then server-side assembly with explainer_video. Use when: "make an explainer video", "explain this in a video", "turn this topic or document into a narrated video", "tell this story as an animated video", "make a faceless narrated video", or "show me explainer styles". Supports live CMS presets, custom style references, mascot/faceless modes, two aspects, and optional burned subtitles. NOT for: photoreal films, ads/UGC, talking heads, podcasts, motion typography reels, one-off clips without narration, or editing a finished video.

Build, edit, and deploy full-stack websites, apps and games via the Higgsfield CLI (`higgsfield website …`). Each is a React 19 + TanStack Start SSR app in one Cloudflare Worker (D1/R2/KV/DO/Containers). THREE product types, picked via `--type` on create: `website` (standalone, no Higgsfield integration — references/website-flow.md), `app` (Sign in with Higgsfield + fnf SDK, Quanta — references/app-flow.md), `game` (realtime multiplayer rooms — references/game-flow.md). Routes to the right flow; each carries its own rules and deploy/publish gates. Use when: "build me a website", "make a landing page", "create a web app", "build a SaaS dashboard / portfolio", "make me a game", "deploy this site", "publish". Also owns GAME ART: "make a spritesheet", "tileable texture", "animate a 3D character", game music/SFX — see the game-* references. NOT for: single image/video/audio generation (higgsfield-generate), product photos (higgsfield-product-photoshoot), marketplace cards (higgsfield-marketplace-cards).

Create job stories using the 'When [situation], I want to [motivation], so I can [outcome]' format with detailed acceptance criteria. Use when writing job stories, creating JTBD-style backlog items, or expressing user situations and motivations.

Install the k-skill bundle, use the unified CLI, resolve credentials, verify the runtime, and optionally configure update checks and GitHub starring.

Route LangSmith work into one workflow packet before touching SDK code. Use when the user needs LangSmith tracing, offline evals, annotation/review queues, prompt-registry decisions, audit/gap review, or cross-service trace propagation for an LLM app or agent workflow. Choose one packet: trace-debug, eval, review, prompt-registry, propagation, or audit. Triggers on: LangSmith, LangChain tracing, `@traceable` / `traceable`, `wrap_openai` / `wrapOpenAI`, datasets, experiments, annotation queues, feedback criteria, Prompt Hub, run trees, trace IDs, or production confidence for an AI feature. Not for generic SLO/alert design, non-LangSmith deployment orchestration, or runtime guardrails outside LangSmith.

Create high-click-through YouTube thumbnails and vertical video covers through the Higgsfield CLI. Builds a truthful information-gap concept, preserves up to three referenced identities, supports logos and controlled variants, renders the main image with Nano Banana Pro, and applies focused Seedream edits. Use when: "make a YouTube thumbnail", "thumbnail for this video", "MrBeast-style cover", "Shorts cover", or "Instagram video cover". Chain after any video workflow once its truthful topic and visual direction are known. NOT for producing the video itself (use higgsfield-generate), product catalog photos (use higgsfield-product-photoshoot), or marketplace cards (use higgsfield-marketplace-cards).

Convert a supplied HTML page or generated HTML reference into a screenshot-backed article containing multiple reusable interaction prompts. Use when the user provides an HTML file, exported page, generated-page.html, or local/live reference and asks to extract animation/interactions, create prompts, capture screenshots for each prompt, add them to an article, or commit the resulting article/assets.

Use kordoc for agent-native HWP/HWPX document parsing, JSON extraction, diffing, form-field extraction, and Markdown→HWPX reverse conversion (read/convert only — for binary editing use rhwp-edit).

Use this skill to benchmark shell commands reliably with warmup runs, statistical summaries, and exportable artifacts using hyperfine.

Identify the Ideal Customer Profile (ICP) from research data with demographics, behaviors, JTBD, and needs. Use when defining your ICP, analyzing PMF survey data, or understanding who your best customers are.

Identify risky assumptions for a feature idea in an existing product across Value, Usability, Viability, and Feasibility. Uses multi-perspective devil's advocate thinking. Use when stress-testing a feature idea, doing risk assessment, or preparing for assumption mapping.

Identify risky assumptions for a new product idea across 8 risk categories including Go-to-Market, Strategy, and Team. Use when evaluating startup risks, assessing a new product concept, or mapping assumptions for a new venture.

Run KADATH (Kernel for Agentic Darwinian Adaptation, Tooling, and Heredity), a Docker-based evolutionary kernel that turns a goal into a locked, Architect-authored benchmark, then evolves a population of smolagents-based coding agents across epochs: each agent runs in an isolated container, gets graded against frozen evidence, and the population is culled, mutated, and reproduced generation over generation until it converges on the best-performing agent framework for that goal. Use when the user wants to propose/approve/run a KADATH evolutionary run, check a run's status or live dashboard, pause/resume/continue a run, export the winning agent population, or understand its Architect/Grader/Tweaker/Birther pipeline, evidence-freezing, or genome lineage/memory model. Triggers on: "kadath", "kadath.sh", "evolve an agent", "Darwinian agent evolution", "agent population fitness benchmark", "smolagents evolutionary run", "kadath dashboard", "genome lineage", "epoch champions".

Search Korean patent and utility-model publications through the official KIPRIS Plus Open API with keyword search plus application-number detail lookup.

Use when the user needs BMAD-style phased delivery for a LangChain, LangGraph, or Deep Agents project and wants the right framework skill at the right BMAD phase. Triggers on: langchain bmad, bmad langchain, langgraph bmad, deep agents bmad, structured agent development, phase-gated langchain workflow, framework-aware bmad, and BMAD for agent frameworks.

INVOKE THIS SKILL when setting up a new project or when asked about package versions, installation, or dependency management for LangChain, LangGraph, LangSmith, or Deep Agents. Covers required packages, minimum versions, environment requirements, versioning best practices, and common community tool packages for both Python and TypeScript.

Create LangChain agents with create_agent, define tools, and use middleware for human-in-the-loop and error handling.

INVOKE THIS SKILL when you need human-in-the-loop approval, custom middleware, or structured output. Covers HumanInTheLoopMiddleware for human approval of dangerous tool calls, creating custom middleware with hooks, Command resume patterns, and structured output with Pydantic/Zod.

INVOKE THIS SKILL when building ANY retrieval-augmented generation (RAG) system. Covers document loaders, RecursiveCharacterTextSplitter, embeddings (OpenAI), and vector stores (Chroma, FAISS, Pinecone).

INVOKE THIS SKILL when writing ANY LangGraph code. Covers StateGraph, state schemas, nodes, edges, Command, Send, invoke, streaming, and error handling.

INVOKE THIS SKILL when implementing human-in-the-loop patterns, pausing for approval, or handling errors in LangGraph. Covers interrupt(), Command(resume=...), approval/validation workflows, and the 4-tier error handling strategy.

INVOKE THIS SKILL when your LangGraph needs to persist state, remember conversations, travel through history, or configure subgraph checkpointer scoping. Covers checkpointers, thread_id, time travel, Store, and subgraph persistence modes.

Define and design a product metrics dashboard with key metrics, data sources, visualization types, and alert thresholds. Use when creating a metrics dashboard, defining KPIs, setting up product analytics, or building a data monitoring plan.

Perform a PESTLE analysis covering Political, Economic, Social, Technological, Legal, and Environmental factors. Use when assessing the macro environment, doing strategic planning, or evaluating external factors affecting your business.

Work with Lapian Notes / 拉片笔记 (github.com/bkingfilm/lapian-notes) — a local- first React/Vite tool that turns a film into an editable shot-by-shot study notebook: local frame extraction, AI-assisted structure analysis (bring your own AI, no API key required), story-line swimlane timeline, structure tree, and audience-emotion curve. Use when the user asks about Lapian Notes, "拉片笔记", "拉片" (shot-by-shot film analysis) tooling, cloning/running this repo (npm run dev, run.bat/run.command), the AI-analysis-package (ZIP) round-trip workflow, or contributing a PR to lapian-notes. Not for generic video editing (use `opencut` for that) or generic film-analysis theory unrelated to this codebase.

Lazyweb is the design-evidence skill for AI coding agents. Use it when designing, critiquing, or changing product UI — it provides real app screenshots, competitor references, best practices, quick examples, creative cross-category ideas, paywall optimization guidance, and mobile growth + monetization A/B test context. Routes to the right Lazyweb mode (design workflow, quick search, update, flowchart, or A/B test research) using MCP tools and the Lazyweb server at https://www.lazyweb.com/mcp. Use before designing any screen when you need design evidence instead of training-data vibes.

Generate a Lean Canvas with problem, solution, metrics, cost structure, UVP, unfair advantage, channels, segments, and revenue. Use when exploring a lean startup canvas, testing a business hypothesis, or modeling a new venture.

Build and maintain a persistent markdown wiki that an LLM updates on the user's behalf, usually inside an Obsidian vault or git-tracked notes repo. Use when raw sources such as web articles, papers, meeting notes, transcripts, screenshots, or past analyses need to be turned into an interlinked knowledge base with immutable source files, LLM-written wiki pages, `index.md`, `log.md`, schema rules in `AGENTS.md` or `CLAUDE.md`, source summaries, query notes, and recurring lint passes. Triggers on: llm-wiki, personal wiki, obsidian wiki, research vault, knowledge base, source ingest, persistent notes, wiki maintenance, source summaries, query filing.

Route runtime-log requests into one evidence packet before diagnosing anything. Use when the user shares app/server/container/browser/CI/JSON log output and wants the first actionable blocker, repeated signature, likely blast radius, or safest next read-only checks. Choose one packet: app-runtime, container-runtime, browser-plus-api, ci-cascade, structured-json, or security-signal. Triggers on: check the logs, which line matters, real error, first blocker, noisy stack trace, retry storm, browser 401/500, pod logs, worker crash, CI abort, webhook failure. Route engine-specific Unity/Unreal logs to `game-build-log-triage`, observability design to `monitoring-observability`, and remediation/debug hypotheses to `debugging`.

Route BigQuery-backed Looker Studio work into one stakeholder-reporting packet: dashboard-spec, slow-dashboard triage, refresh-shape choice, audience split, or exec-handoff. Use when the user needs KPI boards, PM/ops reviews, marketing / GTM reporting, product funnel summaries, or game/live-ops telemetry dashboards on top of curated BigQuery data. Route KPI explanation to `data-analysis`, repeated anomaly hunting to `pattern-detection`, telemetry/alerting coverage to `monitoring-observability`, and semantic-platform choice to `survey`.

Identify 3-5 potential customer segments with demographics, JTBD, and product fit analysis. Use when exploring market segments, identifying target audiences, evaluating new markets, or learning how to segment a market.

Estimate market size using TAM, SAM, and SOM with top-down and bottom-up approaches. Use when sizing a market opportunity, estimating addressable market, preparing for investor pitches, or evaluating market entry.

Routing front door for obsidian-second-brain — a self-rewriting Obsidian vault that evolves Karpathy's LLM-Wiki pattern. Every source REWRITES existing pages instead of appending: people updated, claims revised, contradictions reconciled, patterns synthesized automatically. 45 commands across 4 layers, background + 4 scheduled agents, 4 role presets, AI-first write validator. Cross-CLI: Claude Code, Codex CLI, Gemini CLI, OpenCode.

Route broad product and growth marketing asks into one operating mode, one primary lane, and one reusable operator packet across launch planning, conversion surfaces, lifecycle/retention, acquisition/content, and measurement/experiments. Use when the user needs general marketing help for a website, SaaS product, funnel, onboarding flow, pricing page, campaign, or launch and the first job is deciding the right next brief with owner, dependencies, approvals, and proof instead of producing channel soup. Triggers on: marketing plan, GTM help, launch brief, pricing page refresh, lifecycle email strategy, conversion help, campaign measurement, content plan. Route Steam/store-page game launch work to `steam-store-launch-ops` and backlog/milestone shaping to `task-planning`.

Generate 5 creative, cost-effective marketing ideas with channels, messaging, and engagement rationale. Use when brainstorming marketing campaigns, planning product promotion, or looking for creative marketing tactics.

Build production MCP (Model Context Protocol) servers that let LLMs drive an external API through well-designed tools, in TypeScript (MCP SDK) or Python (FastMCP). Route one request to research and planning, tool design, transport choice, implementation, security review, or evaluation. Use when the user wants to wrap an API as MCP tools, name and shape those tools, choose between stdio and streamable HTTP, decide tools vs resources vs prompts, write a Server Card and publish to the MCP registry, fix a server whose tools the model cannot use correctly, or build the 10-question evaluation that proves it works. Requires confirmation before running an evaluation harness that spends API credits or calls a live service. Owns the retired `mcp-server-design` name. Route consuming an existing MCP server to that server's own skill, and generic API contract design to `api-design`.

Drive the SV Number MCP server (`sv-number-mcp`), nine stdio tools that order a private phone number for one signup, read the SMS verification code straight from the API, and hand the number back — covering 200+ countries. Use when an agent hits a signup form that needs a real phone number for an OTP/SMS code, when picking a country or service code (`list_countries`, `list_services`), ordering and polling a number (`order_number`, `wait_for_code`), closing out an activation (`finish_activation`, `cancel_activation`, `request_another_sms`), checking account balance (`get_balance`), or computing a TOTP/2FA code locally from a shared secret (`totp_code`). Triggers on: "sv-number", "sv-number-mcp", "SMS verification number", "order a phone number for OTP", "receive SMS verification code", "temporary/virtual number for signup", "sms-verification-number.com", "wait_for_code", "totp_code", "phone number MCP server for agents".

Drive mex (`mex-agent`), persistent project memory and code graphs for AI coding agents. One command scaffolds a living wiki, builds a deterministic code graph, and installs a project anchor file (CLAUDE.md, root AGENTS.md, .cursorrules, .windsurfrules, copilot-instructions.md, or .opencode/opencode.json) that your agent auto-loads as a standing rule document. Use when the user wants to `mex setup` a new project, build a symbol-grounded wiki, keep knowledge connected to implementation, route relevant context to agents, or run drift detection (`mex check`, `mex sync`). Triggers on: "mex setup", "project memory", "code graphs", "codebase documentation", "drift detection", "agent memory", "structured scaffolds", "architectural context", "living wiki", "project anchor file".

Design enterprise-grade agent systems with Microsoft's agent framework patterns: role separation, workflow control, policy boundaries, and observability. Use when users need robust organizational agent workflows, governance, and maintainable multi-agent architecture.

Migrate test files from `as` type assertions to @total-typescript/shoehorn. Use when user mentions shoehorn, wants to replace `as` in tests, or needs partial test data.

Drive Mole (`mo`), tw93's GPL-3.0 macOS maintenance CLI that cleans caches and app leftovers, uninstalls apps with their remnants, purges rebuildable project artifacts, removes downloaded installers, explores disk usage, runs bounded system optimization, and reports live health. Routes one request to one mode: run a command safely (`--dry-run` first, the user runs the destructive step), consume the JSON/NDJSON agent surfaces (`mo analyze --json`, `mo status --json` / `--watch`, `mo history --json`, `~/.config/mole/clean-list.txt`), install/update/remove on the right channel, configure whitelists and scan paths, troubleshoot, or contribute to the repo. Use when a user wants to free Mac disk space or fully uninstall a Mac app. Triggers on: mole, `mo clean`, `mo uninstall`, `mo analyze`, `mo purge`, `mo status`, tw93/Mole, mole.fit, clean my Mac, what is eating my disk, CleanMyMac / AppCleaner / DaisyDisk alternative, brew install mole.

Drive Moli (`moli`), Lexmount's open-source headless browser for AI agents, built around on-demand rendering: real JavaScript, DOM, and CSS by default, with layout and pixels computed only when explicitly requested via `--layout`. Use when the user wants to fetch/extract a live JavaScript-rendered page as Markdown/HTML/JSON/semantic-tree, capture a screenshot or PDF, run a small bounded crawl, start a CDP/WebDriver automation server for Playwright/Puppeteer, replace a Chromium/ChromeDriver dependency, or diagnose readiness/network/frame issues on a rendered page. Triggers on: "moli fetch", "moli serve", "headless browser for agents", "on-demand rendering browser", "CDP server without Chrome", "structure-first web scraping", "Lexmount browser", "moli-webfetch", "moli-cdp-server".

Brainstorm 3-5 monetization strategies with audience fit, risks, and validation experiments. Use when exploring revenue models, evaluating pricing strategies, or deciding how to monetize a product.

Route observability work from the current packet into one monitoring brief. Use when the main job is deciding service-health signals, telemetry rollout, dashboard/alert coverage, pipeline freshness/schema monitoring, or game live-ops visibility; choosing between service reliability, telemetry foundation, review audit, data/pipeline, and live-ops modes; and naming one smallest implementation slice. Route outage-log root cause to `log-analysis`, code-level failure isolation to `debugging`, bottleneck tuning to `performance-optimization`, rollout execution to `deployment-automation`, LLM-specific tracing to `langsmith`, and engine-profiler interpretation to `game-performance-profiler`.

Role-level metrics, coordination-failure tests, and transcript grading for multi-agent systems. Use when single-agent eval misses bugs that only appear when agents collaborate — handoff drops, role drift, blame loops, deadlocks.

Transform an output-focused roadmap into an outcome-focused one that communicates strategic intent. Rewrites initiatives as outcome statements reflecting user and business impacts. Use when shifting to outcome roadmaps, making a roadmap more strategic, or rewriting feature lists as outcomes.

Design and review the networking contract for multiplayer games before engine or backend implementation. Use when a team must choose dedicated-authoritative, listen-server, relay-assisted peer, lockstep, snapshot, or asynchronous models; assign authority for movement, combat, inventory, economy, and match state; plan prediction, reconciliation, interpolation, lag handling, matchmaking, reconnects, protocol versions, and cheat resistance; select WebSocket, WebRTC DataChannel, WebTransport, or an engine transport; or build a measured network impairment test matrix. Triggers on: multiplayer game, netcode, authoritative server, replication, rollback, host migration, client prediction, snapshot interpolation, lag compensation, matchmaking, WebSocket, WebRTC, WebTransport.

Build, test, run, and flash NightRun — a bare-metal, no_std Rust UEFI application that boots straight into a local LLM (Llama 3.2, Qwen3, or Granite 4.1) with no operating system underneath. Use when the user wants to build/flash a bootable NightRun USB or Raspberry Pi 5 SD image, convert a GGUF model into the `.nrm` container, run/debug the inference engine on the host or in QEMU/OVMF, or troubleshoot no_std kernel/tokenizer parity issues in the NightRun codebase. Triggers on: "nightrun", "boot into an LLM", "bare-metal LLM runtime", "UEFI LLM appliance", "nrconvert", "nrhost", "cargo xtask", "nrm model file", "flash a bootable LLM USB".

Define a North Star Metric and 3-5 supporting input metrics that form a metrics constellation. Classify the business game (Attention, Transaction, Productivity) and validate against 7 criteria for an effective North Star. Use when choosing a North Star Metric, setting up a metrics framework, learning about the North Star Framework, or deciding what to measure.

Queries managed Google NotebookLM notebooks for citation-backed, source-grounded answers via local Claude Code browser automation; use for uploaded sources, not live web search.

Route Node package-delivery ambiguity into one install packet: temporary Git bridge, SHA-pinned shared bridge, private-auth Git path, tarball / `npm pack` artifact, workspace / `file:` inner-loop, or publish-first registry handoff. Use when the user wants to install an npm / pnpm / Yarn / Bun package from a branch, tag, commit, fork, private repo, monorepo package, or unreleased fix, and the real question is which delivery path is safest rather than how Git or package registries work in general. Triggers on: npm install from GitHub, git dependency, github:owner/repo, git+ssh, git+https, private package from repo, install branch vs commit, monorepo package install, npm pack vs git, and should we publish this instead.

Route obsidian-mind work into the right mode: install/bootstrap the vault (ShardMind vs git clone), the daily standup/dump/wrap-up session loop, capture routes for decisions, incidents, 1:1s, and wins, the performance-graph review flow (/om-review-brief, /om-self-review, /om-peer-scan), vault maintenance and /om-vault-upgrade migration, multi-agent wiring across Claude Code, Codex CLI, and Gemini CLI, and optional QMD semantic search. obsidian-mind is a specific ready-made Obsidian vault template giving coding agents persistent, session-spanning memory through five lifecycle hooks, /om-* commands, subagents, and a competency graph — not a generic vault-building or wiki-authoring workflow. Triggers on: obsidian-mind, om-standup, om-dump, om-wrap-up, om-review-brief, om-self-review, om-peer-scan, om-vault-upgrade, brag doc, North Star.md, performance graph, competency notes, session lifecycle hooks, shardmind install obsidian-mind.

Create, validate, and consume Google's Open Knowledge Format (OKF) bundles — YAML-frontmatter Markdown files with type / title / description / resource / tags / timestamp fields for portable, interoperable AI-agent knowledge sharing. OKF formalizes the LLM-Wiki pattern into a vendor-neutral open specification so any producer can write and any agent can consume without translation. Routes: use `llm-wiki` for raw source capture + vault maintenance, `obsidian` for Obsidian-vault workflows, `graphify` for durable committed graphs, `scrapling` for web-content extraction into OKF docs. Triggers on: okf, open knowledge format, knowledge bundle, okf document, llm wiki standard, knowledge atom, agent context format, okf frontmatter, okf bundle, knowledge interoperability.

Run the Ouroboros specification-first development loop: reduce ambiguity with a Socratic interview grounded in live git data (commits, churn, contributors), freeze an immutable seed/spec, render the execution plan through spec-kit (/speckit.plan → /speckit.tasks), execute against that contract through cli-anything agent-native CLI harnesses (cli-hub, --json evidence), verify before claiming success, and keep looping until completion is actually verified. Use when the user wants spec-first clarification, git-aware interviews, immutable requirements, drift-aware implementation, harness-driven execution, or a persistent completion loop that should keep going until tests / checks / acceptance criteria pass. Triggers on: ooo, ouroboros, interview, seed, run workflow, evaluate, evolve, ooo ralph, specification first, socratic interview, git-aware interview, ambiguity reduction, execution plan, cli harness execute, persistent completion.

Run AI-powered code review through Alibaba's open-code-review (`ocr`) CLI instead of hand-reviewing diffs. Use when the user asks to review code, review a PR/MR, review staged/unstaged/untracked changes, review a single commit, compare two branches, or scan whole files for bugs, security issues, performance problems, and quality concerns. Picks the lightest invocation (workspace review, branch range, single commit, or full-file scan), passes business context via `--background`, classifies findings by priority, and optionally applies fixes. Triggers on: open code review, ocr review, ocr scan, alibaba code review, ai code review cli, review my changes, review this pr, review this commit, review this branch, scan repository for issues.

Route design-generation work into the right Open Design workflow mode — prototype, deck, document, or media artifact. Use when the user needs local-first UI prototype generation, presentation deck creation, or design-system-aware HTML/PDF/PPTX artifacts via locally-installed coding agents. Supports 72 built-in design systems, 5 visual directions, 93 media prompt templates, and multi-format export. Triggers on: open-design, local design tool, prototype generation, design deck, design artifact, design agent workflow, open design prototype, design system artifact, UI prototype generation.

Prepare, generate, and review game UI concepts through the local nexu-io/Open Design app using real runtime screenshots and project assets. Use when Codex must design or redesign Darkbone Archer UI for phone, tablet, desktop, Steam, controller, or multiple locales; create an Open Design handoff/project; drive a local Codex design run; inspect an Open Design HTML artifact; or complete a design-first approval loop before runtime implementation.

Build and validate Open Design handoffs from the real Darkbone Archer runtime — game UI (upgrade-only preservation contracts) AND VFX/skill/RIG design packages. Use whenever Codex must prepare, regenerate, or audit an Open Design handoff; feed an existing Claude Design handoff zip into Open Design; collect current screenshots, animation videos, keyframes, source anchors, assets, responsive states, or preservation boundaries; or prevent a standalone concept from deleting existing identity, motion, interaction, copy, or data that it does not show.

Brainstorm product positioning ideas differentiated from competitors. Identifies top competitors and generates positioning statements with rationale. Use when developing product positioning, differentiating from competitors, or crafting brand positioning strategy.

Run a pre-mortem risk analysis on a PRD or launch plan. Categorizes risks as Tigers (real problems), Paper Tigers (overblown concerns), and Elephants (unspoken worries), then classifies as launch-blocking, fast-follow, or track. Use when preparing for launch, stress-testing a product plan, or identifying what could go wrong.

Take over an approved Open Design artifact (game UI, VFX, skill, RIG, enemy/boss) into Darkbone Archer without degrading the runtime. Use whenever Codex must analyze, port, integrate, or review an Open Design standalone; apply a responsive Home, Map Select, Talent, Mask, Fusion, Victory, or settlement design; port an Open Design VFX/RIG lab into Phaser; preserve existing animations omitted by a mockup; or prove current-runtime vs standalone vs integrated-runtime fidelity before merge.

Operate Anil-matcha/Open-Generative-AI, the MIT-licensed Next.js and Electron studio that fronts MuAPI image, video, audio, lip sync, workflow, and agent models with optional local sd.cpp and Wan2GP inference. Route one request to fit check, desktop release install, source or Docker self-host, API key and provider configuration, local inference setup, model catalog verification, troubleshooting, or upgrade. Use when the user names Open Generative AI, Anil-matcha/Open-Generative-AI, its Image, Video, Cinema, Lip Sync, Workflow, or Design Agent studios, `npm run electron:dev`, MuAPI keys, sd.cpp, or Wan2GP. Require confirmation before installer execution, unsigned-binary Gatekeeper or SmartScreen overrides, AppArmor sysctl changes, credential entry, multi-gigabyte model downloads, public deployment, or paid generation. Route provider-neutral programmatic video pipelines to `video-production` and local desktop video editing to `opencut`.

Adapts the pinned Open-Generative-AI model catalog to a safe shopping-shorts render plan, choosing MuAPI, GUI handoff, or Wan2GP while enforcing 9:16 support, product identity, provenance, cost review, and no-secret logging.

Build and operate multi-agent workflows with OpenAI Agents SDK (Python): define agents/tools/handoffs, add guardrails, run conversations, and debug orchestration behavior. Use when users ask for agent orchestration with OpenAI-native patterns, handoff routing, or production-ready agent loops.

Route active project/repo memory requests into one honest packet: memory-layer choice, load-context, search-context, store-conclusions, setup-integration, or repo-packer route-out. Use when agents need searchable decisions, manifests, stable links, handoff notes, and small “read this first” packets across sessions. Route long-lived markdown knowledge bases to `llm-wiki`, structural graph memory to `graphify`, human-authored vault organization to note/vault skills, and one-shot repo packing to tools like Repomix, Gitingest, or Code2Prompt.

Generate SQL queries from natural language descriptions. Supports BigQuery, PostgreSQL, MySQL, and other dialects. Reads database schemas from uploaded diagrams or documentation. Use when writing SQL, building data reports, exploring databases, or translating business questions into queries.

Operate Browserbase Stagehand, the MIT-licensed browser-agent SDK and `browse` CLI for local or Browserbase Chromium automation, natural-language `act`, `observe`, and `extract`, typed WebMCP calls, cloud Search/Fetch, Playwright migrations, and Browserbase MCP integration. Use when a user wants to build, run, debug, or migrate a Stagehand browser agent, choose local versus cloud browser execution, or configure the `browse` command surface. Triggers on: stagehand, @browserbasehq/stagehand, browserbase.launch, localBrowser.launch, act(), observe(), extract(), WebMCP, browse CLI, Browserbase MCP.

Build a stakeholder map using a power/interest grid, identify communication strategies per quadrant, and generate a communication plan. Use when managing stakeholders, preparing for a launch, aligning cross-functional teams, or planning stakeholder engagement.

Turn messy daily-sync requests into one honest coordination-cadence packet: decide whether live-daily, async-daily, hybrid-daily, less-frequent syncs, or no recurring standup is justified, then pick the lightest standup mode that still catches blockers and handoffs. Use when the user needs a daily scrum, async check-in, board-walk, blocker-first sync, remote-team standup, launch- week cadence reset, or help fixing an overlong / low-signal standup. Route backlog planning to `task-planning`, sizing to `task-estimation`, retrospective process repair to `sprint-retrospective`, and incident-command work to debugging / launch-specific skills.

Work with the OpenCut open-source video editor repo (github.com/OpenCut-app/OpenCut) — clone/setup, run the web/desktop dev servers, understand the Rust/WASM compositor core, and follow current contribution focus areas. Use when the user asks about OpenCut, the "open-source CapCut alternative", opencut-classic, setting up the OpenCut monorepo (moon/proto or Bun/Docker), the OpenCut Rust core, or contributing a PR to OpenCut. Also triggers on "opencut.app", "new.opencut.app", and "opencut-app/opencut-classic".

Operate SenteLabsAI/OpenExecutive, the Apache-2.0 self-hosted virtual executive team that answers business questions through one Executive persona backed by eight specialist Claude agents over FastAPI, Next.js, ChromaDB, and SQLite. Route one request to one mode: fit-check the project; preflight Python 3.11, uv, and Node 22; choose an Anthropic, OpenRouter, or local model provider; run it with `make dev` or Docker; control paid spend and Anthropic prompt caching; connect Slack, Discord, Telegram, Google Chat, or Gmail without sending unwanted messages; operate fixtures, the single-instance scheduler, and Fly.io deploys; or contribute an agent with tests and evals. Use when a user wants to run, configure, extend, or debug Open Executive. Triggers on: openexecutive, Open Executive, SenteLabs, virtual executive team, virtual CFO or CSO agent, consult_specialist, openexec-api, episodic memory executive, fixtures reset, single-instance scheduler.

Operate calesthio/OpenMontage, the AGPL-3.0 agent-orchestrated video production repository where a coding agent reads YAML pipelines and Markdown directors, invokes Python tools, checkpoints JSON artifacts, and exposes runs in the Backlot board. Route one request to one mode: assess and bootstrap a checkout; select a live pipeline from `pipeline_defs` and run `provider_menu_summary()` preflight; analyze a reference video; start or resume a checkpointed production with cost and human gates; inspect Backlot/project state; or extend and verify a provider, pipeline, renderer, or contract. Use when the user names OpenMontage or is operating its repository. Triggers on: OpenMontage, calesthio/OpenMontage, AGENT_GUIDE.md, pipeline_defs, provider_menu_summary, decision_log, Backlot, video_compose, Remotion vs HyperFrames, checkpoint_<stage>.json, agentic video production.

Operate openocta/openocta, the Apache-2.0 desktop and service AIOps agent for natural-language inspection, alert analysis, data queries, remediation, local knowledge, Skills, MCP, channels, and webhooks. Route one request to fit check, release install or upgrade, gateway and model configuration, security hardening, integration, troubleshooting, or source build. Use when the user names OpenOcta, Open Octa, its `openocta` CLI, port 18900, Knowledge Vault, OpenOcta Skills, digital employees, or the openocta/openocta repository. Require confirmation before installer execution, credentials, service or channel activation, marketplace installs, production remediation, or uninstall. Route generic observability design to `monitoring-observability` and raw log triage to `log-analysis`.

Install and route through OpenSpace, the skill management layer for AI agents, so a host agent can retrieve/rank/load the right SKILL.md out of this jeo-skills catalog (~150 installed skills), then evaluate skill quality from real execution evidence and evolve skills via FIX/DERIVED/CAPTURED updates. Covers install-as-skill-finder, retrieve-a-skill, evaluate-quality, evolve-skills, and local-first hub share/import. Triggers on: openspace, skill finder, skill retrieval, find the right skill, rank skills, skill discovery, skill quality, evolve skill, FIX DERIVED CAPTURED, skill hub, openspace-mcp, DiscoverSkills, skill-discovery, delegate-task.

AI 애니메이션 스프라이트 생성. 텍스트 설명 한 줄로 캐릭터 + 동작 애니메이션(걷기·달리기·공격·마법 등 100여 종) + 8방향 스프라이트 세트를 만들고, 게임 엔진용 번들(스프라이트시트 · manifest.json · Aseprite JSON · 상태별 GIF/APNG · 개별 프레임 PNG)로 내보냅니다. Use when the user wants to generate game sprites, character animations, sprite sheets, sprite atlases, or 8-direction sprite sets from a text description. Triggers on: perfectpixel, ppgen, sprite generation, character animation, sprite sheet, sprite atlas, 8-direction sprite, god-tibo-imagen sprite, gemini sprite.

Route performance work from the artifact people already have into one measurement-led tuning brief. Use when the main job is locating the tightest latency, throughput, memory, bundle-size, or frame-budget bottleneck; choosing the right trace, query plan, load-test result, profiler capture, or CWV report; naming one bottleneck; and recommending one or two bounded optimizations with before/after verification. Route telemetry rollout to monitoring-observability, correctness diagnosis to debugging, structural cleanup to code-refactoring, validation-policy design to testing-strategies, and engine-specific capture interpretation to game-performance-profiler.

Work with OpenStory (github.com/openstory-so/openstory), the open-source AI script-to-video sequence platform built on Bun + TanStack Start + Cloudflare Workers (D1, R2, Workflows, Durable Objects) with Drizzle, Better Auth, and Fal.ai/OpenRouter models. Routes one request to one mode: run it locally (`bun install && bun dev` on Miniflare, `FAL_KEY` / `OPENROUTER_KEY`), trace the storyboard pipeline (scene split, casting, prompts, frame images, motion, music), author a Cloudflare Workflow under the three-place wiring and no-mid-run-read `scopedDb` contract, add or update an image/video/audio/LLM model, ship a safe D1 + Drizzle migration, or deploy. Use when the user mentions OpenStory or is operating this codebase. Triggers on: openstory, openstory.so, script to video, AI video sequence platform, storyboard workflow, triggerWorkflow, OpenStoryWorkflowEntrypoint, WorkflowScopedDb, IMAGE_MODELS, fal.ai model registry, D1 CASCADE data loss, flatten-migrations, deploy Cloudflare Workers video app.

Run Comet's Opik — open-source LLM observability, evaluation, and optimization — from one routing-first skill: install the Python/TypeScript SDK, stand up a server (Comet.com cloud, Docker Compose via `./opik.sh`, or Kubernetes/Helm), wire tracing through `@opik.track` or one of 50+ framework integrations (OpenAI, Anthropic, LangChain, LangGraph, LlamaIndex, CrewAI, DSPy, Ollama, Bedrock, Vercel AI SDK, …), score outputs with LLM-as-a-judge metrics (Hallucination, Moderation, Answer Relevance, Context Precision), and run Datasets/Experiments evaluations including PyTest CI gates. Use when the user wants LLM tracing, prompt evaluation, production LLM monitoring, agent optimization, or guardrails with Opik. Triggers on: opik, comet opik, opik configure, opik.sh, llm observability, llm tracing, llm as a judge, hallucination metric, prompt evaluation, opik dashboard, opik guardrails, agent optimizer.

Build an Opportunity Solution Tree (OST) to structure product discovery — map a desired outcome to opportunities, solutions, and experiments. Based on Teresa Torres' Continuous Discovery Habits. Use when structuring discovery work, mapping opportunities to solutions, or deciding what to build next.

Profile, audit, and optimize frontend page performance with emphasis on animation work, memory-leak risks, long-session slowdowns, CSS animations, canvas/WebGL requestAnimationFrame loops, marquees, skeletons, GSAP/Three/Matter effects, timers, listeners, and observers. Use when the user asks to make animations performant, pause offscreen animations, look for memory leaks, profile pages that slow the computer over time, fix janky scrolling, reduce CPU/GPU use, or repeat the "only play in view" optimization on React/Vite/Next/frontend pages using Codex Browser.

Drive Palmier Pro, an open source AI-native macOS video editor (Swift, SwiftUI/AppKit, AVFoundation) that exposes its timeline as an MCP server at `http://127.0.0.1:19789/mcp` so Claude Code/Desktop, Cursor, or Codex can read and edit a project's tracks, clips, media, transcript, captions, color/effects, and trigger generative AI (video/image/audio) requests side-by-side with a human editor. Use when the user wants to connect an agent to Palmier Pro's MCP server, call its timeline/clip/media/generation tools (`get_timeline`, `add_clips`, `move_clips`, `generate_video`, ...), build/run/test the Swift app from source, or debug the MCP tool surface in `ToolDefinitions.swift`/`ToolExecutor+*.swift`. Triggers on: "palmier pro", "palmier-pro", "AI video editor MCP", "connect Claude to my video editor", "palmier MCP server", "edit my timeline with an agent", "swift build PalmierPro", "palmier-pro mcpb", "manage_project"/"get_timeline"/"add_clips" tool.

Route academic-figure work into the lightest workable PaperBanana mode instead of jumping straight to a full multi-agent generation run. Use when the user needs publication-quality methodology diagrams, statistical plots, figure evaluation, polishing an existing figure, batch/sweep generation, or a full-paper figure package from text or PDF. Even if the user does not say "paperbanana" — also triggers on: academic figure, methodology diagram, publication figure, generate diagram from paper, statistical plot from CSV, figure evaluation, polish figure, NeurIPS/ICML figure, arxiv illustration, plan-then-refine diagram pipeline.

Route repeated pattern, rule, and anomaly work into one detection packet before suggesting tools or fixes. Use when the user needs reusable scans, suspicious repeated shapes, grouped outlier candidates, or first-pass anomaly triage across code, logs/events, telemetry, and metric tables. Choose one packet: text-prefilter, structural-code-rule, log-event-pattern, or metric-anomaly. Triggers on: repeated bug, suspicious spike, odd cohort, noisy event, code smell family, rule pack, anti-pattern, outlier, anomaly, fraud signal, or recurring issue. Route root-cause incident work to log-analysis, KPI/business explanation to data-analysis, repo tracing to codebase-search, remediation to specialist skills, and alert/incident operations to monitoring-observability.

Operate Payload CMS (Next.js-native headless CMS) in repo workflows: bootstrap a Payload app, configure collections/globals, run local dev + migrations, and ship safe content-model changes. Use when the request mentions Payload CMS, payload config, collection schema, admin panel, or Next.js + headless CMS integration.

Routing-first visual approval gate for AI agent plans, markdown specs, and diffs. Use when a human needs to review a concrete plan before execution, inspect a targeted diff in a browser, mark up a spec/PRD/architecture note, or set up the review loop on Claude Code, Gemini CLI, Codex CLI, or OpenCode. Route planning/spec creation to `task-planning` or `ralph`, broad PR-policy review to `code-review`, rendered-UI critique to `agentation`, and fresh-session browser verification to `browser-harness`.

Reuse a running Chrome session for browser automation via Playwriter CLI + MCP. Use when the task depends on the browser the user already has open — existing logins, cookies, extensions, passkey-friendly flows, or live-tab continuity. Route repeatable fresh-browser or CI-style checks to `browser-harness` instead.

AI operating system for product managers — 65 skills, 36 commands across 8 plugins encoding PM frameworks by Teresa Torres, Marty Cagan, Alberto Savoia, and other PM thought leaders. Use when doing product discovery, writing PRDs, defining strategy, conducting user research, planning go-to-market, running data analytics, or managing sprints. Triggers on: pm-skills, product manager, product discovery, write PRD, user story mapping, product strategy, north star metric, opportunity solution tree, sprint planning, go-to-market, market research, product roadmap, pretotyping, jobs to be done.

Make the agent solve coding tasks with the least code that remains correct. Before writing code, walk the Ponytail ladder: skip what need not exist, then prefer stdlib, native platform features, already-installed dependencies, one line, and only then the minimum custom code. Use when the user asks for ponytail mode, less code, YAGNI, anti-bloat, minimal code, an over-engineering review, a current-diff delete-list, a whole-repo bloat audit, or a `ponytail:` tech-debt harvest. Keep validation, data-loss handling, security, and accessibility. Mark shortcuts with `ponytail:` plus the upgrade path. Triggers on: ponytail, /ponytail, /ponytail-review, /ponytail-audit, /ponytail-debt, write less code, YAGNI, over-engineering, anti-bloat, minimal code, do I need this, lazy dev.

Perform Porter's Five Forces analysis — competitive rivalry, supplier power, buyer power, threat of substitutes, and threat of new entrants. Use when analyzing industry dynamics, assessing competitive forces, or evaluating market attractiveness.

Build real deck artifacts when the user needs editable slides, not just prose: investor decks, roadmap/QBR decks, launch decks, architecture/demo decks, workshop/training decks, and game pitch or milestone decks. Use when the job is to choose one deck mode, one smallest useful artifact packet, and one honest handoff surface (HTML review, PPTX, PDF, Google Slides, or Figma Slides). Triggers on: presentation, slide deck, slides, pitch deck, roadmap deck, investor deck, launch deck, architecture review deck, demo deck, workshop slides, keynote, board deck, QBR deck, and game pitch deck. Route long-form docs to `technical-writing`, end-user tutorials to `technical-writing`, research manuscripts to `research-paper-writing`, and broad marketing planning to `marketing-automation`.

Fast, accurate, and comprehensive multiline text measurement and layout library for JavaScript/TypeScript — without DOM reflow. Use when the user needs to calculate paragraph heights, line counts, manual line layouts (text flowing around floated images), or rich text rendering (emoji, CJK, RTL) on DOM, Canvas, or SVG targets. Covers npm install, core API (prepare/layout, prepareWithSegments/layoutWithLines), and options (whiteSpace, wordBreak, letterSpacing). Triggers on: pretext, text measurement, text layout, paragraph height, line layout, text reflow, canvas text measurement, DOM reflow.

Analyze and design pricing strategies including pricing models, competitive pricing analysis, willingness-to-pay estimation, and price elasticity. Use when setting prices, evaluating pricing models, preparing for a pricing change, or comparing freemium vs paid approaches.

Reference guide to 9 prioritization frameworks with formulas, when-to-use guidance, and templates — RICE, ICE, Kano, MoSCoW, Opportunity Score, and more. Use when selecting a prioritization method, comparing frameworks like RICE vs ICE, or learning how different prioritization approaches work.

Prioritize assumptions using an Impact × Risk matrix and suggest experiments for each. Use when triaging a list of assumptions, deciding what to test first, or applying the assumption prioritization canvas.

Prioritize a backlog of feature ideas based on impact, effort, risk, and strategic alignment with top 5 recommendations. Use when prioritizing a feature backlog, making scope decisions, or ranking product ideas.

Draft a detailed privacy policy covering data types, jurisdiction, GDPR and compliance considerations, and clauses needing legal review. Use when creating a privacy policy, updating data protection documentation, or preparing for compliance.

Brainstorm 5 unique, memorable product names with rationale aligned to brand values and target audience. Use when naming a new product, rebranding, or exploring product name ideas.

Create a comprehensive product strategy using the 9-section Product Strategy Canvas — vision, segments, costs, value propositions, trade-offs, metrics, growth, capabilities, and defensibility. Use when building a product strategy, creating a strategic plan, or defining product direction.

Brainstorm an inspiring, achievable, and emotional product vision that motivates teams and aligns stakeholders. Use when defining or refining a product vision, creating a vision statement, or aligning the team around a shared direction.

Generate a Startup Canvas combining Product Strategy (9 sections) and Business Model (costs + revenue) for a new product. An alternative to BMC and Lean Canvas that separates strategy from business model. Use when launching a new product or evaluating a startup concept.

Discover and apply curated prompts from the prompts.chat collection to optimize AI interactions. Use when refining prompt engineering, finding domain-specific prompt templates, improving response quality, or building prompt-based workflows. Triggers on: prompt optimization, prompt templates, prompt engineering, prompt library, curated prompts, prompt discovery, and AI prompt patterns.

Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

Build typed LLM applications with PydanticAI: schema-constrained outputs, tool integration, validation, retries, and deterministic downstream handoffs. Use when users need reliable structured outputs instead of free-form text generation.

Run measurement-led React and Next.js performance audits for waterfalls, bundle size, RSC/server-client boundaries, hydration mismatch, rerender churn, script cost, and slow page interactions. Use when the user needs help diagnosing or refactoring a slow React UI, App Router route, heavy client component, or Next.js page with excess JavaScript or weak client/server boundaries. Triggers on: React perf, Next.js perf, waterfall, bundle size, hydration, rerender, client component too heavy, slow route, React profiler, web vitals.

Integrate, customize, and contribute to react-bits (github.com/DavidHDev/react-bits) — the largest and most creative library of animated React components. It covers Vite, Tailwind CSS v4, Three.js/Fiber, GSAP, and Framer Motion integrations. Use when the user asks about react-bits, animated React components, jsrepo registry, or contributing new animated components to the react-bits repository.

Capture exact React element context from a live browser UI and hand it to an AI coding agent as component name, source file path, line number, and HTML. Use when the user wants `react-grab`, element-context copy, component-source lookup from the browser, clipboard-to-agent React debugging, or MCP-backed element selection for React apps. Not for generic browser automation or login/session reuse (`browser-harness`, `playwriter`), broad UI annotation/review (`agentation`), React performance audits (`react-best-practices`), or general design-system work (`design-system`). Triggers on: react-grab, grab element context, copy component to AI, browser component picker, React component inspector, clipboard component source, get element context from browser, grab UI element.

Critic node with verbal-memory of past failures, stable rubric scoring, and bounded retry budget. Use when an agent loop fails repeatedly on the same task and naive retry gives no improvement — Reflexion turns failure into actionable verbal feedback.

Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.

Move issues and external PRs through a state machine of triage roles, categorise, verify, grill if needed, and write agent-ready briefs. Use when issues or external PRs need categorising, verifying, and turning into agent-ready briefs. Triggers on: triage, triage the backlog, review these issues.

Install and drive Motion Previs Studio v4 from the `video-motion-previs` CLI to turn reference video into pose, depth, camera-motion, control-layer, and AI-video/Blender production packs. Use when an agent must check or install the desktop app, import and trim a shot, run motion analysis, export a bundle, inspect its files, capture the app, or send a layer to Blockout. Triggers on: Motion Previs Studio, video motion previs, camera solve, pose extraction, OpenPose BODY_25, depth control video, production pack, motion-previs MCP.

Draft and revise ML, CV, NLP, and systems research papers with strong claim-evidence flow, reviewer-aware structure, and submission-ready section planning. Use when the user needs help with an abstract, introduction, related work, method, experiments, ablations, discussion, figures/tables, checklist coverage, or rebuttal / response-letter writing for a paper. Triggers on: research paper, academic paper, ML paper, NeurIPS paper, ICLR paper, CVPR paper, experiments section, ablation plan, related work, contribution framing, reviewer response, rebuttal, camera-ready revision.

Use when you need to resolve an in-progress git merge/rebase conflict.

Routing-first responsive layout strategy and verification for web interfaces. Use when the main job is classifying whether the failure is page-shell adaptation, reusable component/container behavior, dense-data or toolbar pressure, responsive media, or reflow verification — then turning vague “breaks on mobile” requests into one concrete strategy packet. Route component API design and system-wide breakpoint/token governance to `design-system`, and accessibility remediation plus broad UI audit work to `web-accessibility`.

Comprehensive PM resume review and tailoring against 10 best practices including XYZ+S formula, keyword optimization, job-specific tailoring, and structure. Use when reviewing a PM resume, preparing for job applications, or improving resume impact.

Generate, convert, preview, and integrate retro game sound effects with raysan5/rfxgen. Use when a user needs coin, laser, explosion, power-up, hit, jump, or blip SFX; wants `.rfx` parameters converted to WAV/raw/C headers; needs batch CLI generation; or must build and troubleshoot rFXGen. Triggers on: rFXGen, rfxgen, sfxr, chiptune SFX, .rfx, procedural game sound, or sound preset.

Debug HWP layout, dump document IR, compare versions, extract thumbnails, and unlock read-only HWPs with the upstream rhwp Rust CLI (export-svg/dump/dump-pages/ir-diff/thumbnail/convert).

Edit HWP documents — insert/delete text, replace-all, create tables, set cell text — with the k-skill-rhwp CLI that wraps the @rhwp/core WASM engine (rhwp by Edward Kim).

Install, initialize, verify, and troubleshoot RTK (Rust Token Killer) for AI coding agents. Use when you need to reduce shell-command token output, confirm that the correct `rtk` binary is installed, choose between Homebrew, install.sh, or Cargo installation, wire `rtk init` for Claude Code, Codex, Gemini CLI, Cursor, Copilot, Windsurf, Cline, or OpenCode, or use compact wrappers such as `rtk git status`, `rtk read`, `rtk grep`, `rtk test`, `rtk lint`, and `rtk gain`. Triggers on: rtk, rust token killer, token saver cli, rtk init, rtk gain, codex rtk, gemini rtk, opencode rtk, claude hook token reduction.

Create exercise directory structures with sections, problems, solutions, and explainers that pass linting. Use when user wants to scaffold exercises, create exercise stubs, or set up a new course section.

Audit, inventory, route, selectively install, and safely refresh the K-Dense-AI/scientific-agent-skills collection for scientific packages, databases, lab integrations, research methods, and publication workflows. Use when the user names Scientific Agent Skills, K-Dense scientific skills, or that repository; needs the correct upstream sub-skill; wants a pinned subset installed or refreshed; or needs provenance, license, collision, and security review before adoption. Inspect the real skill tree and each selected license. Never wholesale-vendor the pack, and do not copy or adapt its proprietary Anthropic-derived docx, pdf, pptx, or xlsx folders. Require separate approval before dependency installation, credential use, paid APIs, cloud jobs, lab hardware, clinical outputs, or publication. Route ordinary paper pipelines to `academic-research`, general web research to `deep-research`, figures to `paperbanana`, and scientific LLM evaluation to `scientific-llm-benchmarks`.

A comprehensive reference of benchmarks for evaluating large language models on scientific reasoning and discovery.

Fetch live web pages for agents through ScrapingAnt's hosted MCP server (`https://api.scrapingant.com/mcp`) with headless-Chrome rendering, rotating datacenter/residential proxies, Cloudflare and anti-bot handling, and LLM-ready Markdown output. Use when a plain fetch/WebFetch is blocked (403/429, Cloudflare challenge), when a JavaScript/SPA page returns an empty shell, when geo-specific content is needed, or when standing up a local browser scraper costs more than the task is worth. Triggers on: scrapingant, MCP web scraping, fetch blocked page, Cloudflare bypass, anti-bot scraping, JS rendered page, scrape to markdown, residential proxy fetch, geo-targeted scrape, live web access for agents.

Route web-scraping work into the lightest workable Scrapling mode instead of defaulting to a browser. Use when the user needs HTML extraction, JS-rendered page retrieval, protected-target escalation, quick CLI scraping, agent-facing MCP access, or a larger crawl with Scrapling spiders. Triggers on: scrapling, scrape website, crawl site, adaptive scraping, selector drift, stealthy fetch, browser scraping, scrape to markdown, scrapling mcp, scrapling spider, research harvesting, literature scraping, paper metadata.

Routing-first skill for web/application/API hardening. Use when the main job is classifying which security layer is missing — browser/perimeter policy, session/cookie/CSRF, abuse controls, validation/unsafe execution, secrets/runtime config, or verification — and turning vague OWASP/security asks into one concrete hardening brief. Route auth-stack choice to `authentication-setup`, schema work to `database-schema-design`, code-level bug fixing to `debugging` / `code-review`, and environment wiring to `system-environment-setup`.

Fast, accurate code search for AI agents using ~98% fewer tokens than grep+read. Indexes any local or remote repository in under a second (~250ms on CPU, no GPU or API key needed). Supports natural-language and symbol queries, semantic similar-code discovery, and MCP server integration for Claude Code, Codex, Cursor, and OpenCode. Python library available for programmatic use. Triggers on: semble, code search, semantic code search, semble search, token-efficient search, find code, code search mcp, agent code search, semble find-related, semble savings.

Analyze user feedback data to identify segments with sentiment scores, JTBD, and product satisfaction insights. Use when analyzing user feedback at scale, running sentiment analysis on reviews or surveys, or identifying satisfaction patterns.

Configure this repo for the engineering skills: set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills. Use when this repo has not yet been configured for the suite. Triggers on: setup-matt-pocock- skills, configure the issue tracker, set up triage labels.

상품 링크나 권리 확인된 제품 사진을 근거 중심의 9:16 쇼핑 쇼츠 패키지로 전환한다. Use when 쇼핑 쇼츠·상품 추천 숏폼·어필리에이트 영상·AI 제품 광고를 조사, 기획, 생성, 검수하거나 YouTube·Instagram Reels·TikTok·X·네이버 클립과 YPP·쿠팡·지그재그 등 한국 크리에이터 수익 경로를 현재 공식 조건으로 비교할 때. Open-Generative-AI를 영상 생성 표면으로 연결하되 가짜 체험담, 경쟁 영상 복제, 대량 양산, 무고지 제휴, 승인 없는 가입·유료 생성·업로드에는 사용하지 않는다.

Orchestrates one evidence-backed shopping Short from current Korean platform revenue research through route eligibility, product evidence, rights, concept approval, Open-Generative-AI rendering, ffmpeg assembly, independent QA, and confirmation-gated release.

Route reusable skill-improvement work into one bounded repo-local ratcheting packet: ratchet eligibility, benchmark readiness, loop charter freeze, baseline scoring, one-change mutation, support-surface sync, or final keep/revert report. Use when an existing `SKILL.md`, SOP, prompt, or workflow doc may need sharper triggers, clearer instructions, better support files, or cleaner discovery wording and you want a frozen evaluation harness, append-only experiment logs, and explicit keep-or-revert decisions instead of ad hoc rewriting. Also use when you need to prove that no ratchet is justified yet. Not for GPU-bound Karpathy `autoresearch` runs or hosted app-scale eval / observability platforms such as LangSmith, Braintrust, Weave, or Promptfoo.

Standardize and validate SKILL.md files against the Agent Skills specification (agentskills.io). Use when creating or rewriting a skill, auditing an existing skill for spec compliance, sharpening trigger descriptions, canonicalizing overlapping skills into a canonical skill plus compatibility alias, or checking whether derived discovery surfaces (`skills.json`, README/setup inventories, `SKILL.toon`, `SKILL.compact.md`) still match the live skill folders. In repo-root maintenance loops, prefer truthful validator commands instead of bare `scripts/...` examples. Triggers on: "validate skill", "create SKILL.md", "check skill spec", "skill frontmatter", "improve skill description", "catalog sync", "SKILL.toon drift", "compact skill drift", "canonical skill".

Generate, visually edit, and export beautiful HTML/CSS presentation decks with agents using slides-grab (NomaDamas, MIT) — the open-source Claude Design alternative and best harness + editor + linter for slides in Claude Code / Codex. One routing-first skill across Plan (agent drafts an outline), Design (each slide is a self-contained slide-XX.html), Edit (a pure-JS browser editor where you drag a bbox over any region and ask the agent to rewrite just that area, or hand-tweak text/size/bold), and Export (capture-or-print PDF, per-slide PNG incl. Instagram 1:1 card-news, plus experimental/unstable PPTX and Figma-importable PPTX). Picks an install path (npm package + npx skills add, or clone), a deck workspace (--slides-dir, multi-deck decks/<name>/), one of 35 bundled design styles, and the supported asset flow (local ./assets/<file> only — image via god-tibo-imagen/codex/nano-banana, fetch-video via yt-dlp, tldraw .tldr→SVG), validating with slides-grab validate before any export.

Audit, selectively adapt, install, route, and maintain the Korean solopreneur skill pack from bam-bam-2/solo-skills. Use when the user names solo-skills, bam-bam-2/solo-skills, its 26-skill collection, or `fleet.md`; wants a pinned inventory or selective install; needs to choose among its content, messaging, publishing, remote, media, or agent workflows; or needs to remove author-specific assumptions before adoption. Inspect frontmatter, support scripts, licenses, destination collisions, personal paths, account IDs, credential names, network targets, schedulers, permission-bypass flags, and live-action switches before copying or running anything. Require explicit approval for external messages, publishing, archives, remote commands, desktop control, scheduled agents, paid providers, account access, or local persistence. Route generic agent-team design to the canonical local `harness` and reusable skill authoring to `skill-standardization`.

Drive Soup (`soup-cli`), a CLI-first tool for fine-tuning and post-training LLMs with one YAML config and one command — SFT, DPO/GRPO/ORPO/SimPO/KTO, QLoRA/DoRA/LoRA+, layer streaming for 4-8 GB GPUs, eval-gated training, and serving. Use when the user wants to `soup init`/`soup train` a model, pick a training method or quantization scheme, estimate cost/memory before training, fine-tune on a small local GPU, migrate a config from Axolotl/LLaMA-Factory/Unsloth, or serve/merge/push a trained adapter. Triggers on: "soup-cli", "soup train", "soup init", "fine-tune an LLM locally", "QLoRA on a laptop GPU", "layer streaming", "soup advise", "soup autopilot", "DPO/GRPO/ORPO training", "merge LoRA adapter".

Run GitHub's Spec-Driven Development (SDD) workflow via the `specify` CLI — install spec-kit, initialize a project for one of 30+ AI coding agents (Claude Code, Copilot, Gemini, Cursor, Codex, Qwen, opencode, Kiro, etc.), and drive the constitution → specify → plan → tasks → implement command pipeline. Use when the user wants to bootstrap a Spec-Driven Development project, install `specify-cli`, generate executable specs before code, or invoke the `/speckit.*` slash commands (`/speckit.constitution`, `/speckit.specify`, `/speckit.plan`, `/speckit.tasks`, `/speckit.implement`, `/speckit.clarify`, `/speckit.analyze`, `/speckit.checklist`). Triggers on: spec-kit, speckit, specify, specify init, spec-driven, spec driven development, SDD, /speckit, executable spec.

Facilitate sprint retrospectives, milestone postmortems, or iteration reviews that turn completed work into a few owned process improvements instead of another stale template ritual. Use when the user needs a retrospective mode, remote or hybrid facilitation plan, action-item follow-through reset, or help reviewing what the team should change after a sprint, release, milestone, or rough delivery cycle. Also owns the retired `retro` name: run a retro, start/stop/continue, 4Ls, sailboat, action items with owners and deadlines, compact sprint summary with velocity vs commitment and carry-over. Route backlog planning to `task-planning`, sizing to `task-estimation`, daily coordination to `standup-meeting`, and deep incident forensics to debugging/incident-specific skills.

Choose the smallest viable React/fullstack state owner before comparing libraries. Use when the user needs to classify local UI state, shared subtree/Context state, URL or form state, server-state caches, or client workflow stores such as Zustand / Redux Toolkit / Jotai. Triggers on: global state, prop drilling, Context vs Zustand, Redux Toolkit vs Zustand, server state vs client state, React Router state management, optimistic updates, URL state, form state, and too much state in one store.

Turn Steam store-page, wishlist, demo, Next Fest, and launch-window ambiguity into one packet-first Steam launch brief. Use when an indie dev, small studio, founder-marketer, or publisher helper needs to decide whether the next move is a page-promise audit, wishlist-signal check, demo-readiness gate, event-timing workback, or launch-ops runbook — especially when they say "help my Steam page", "wishlists are weak", "is our demo ready", "should we do Next Fest", or "give me a Steam launch checklist". Route broad non-game GTM work to `marketing-automation` and player-feedback/build-performance issues to the game specialist skills.

A collection of Agent Skills for the Stitch MCP server: generate high-fidelity UI screens, create multi-page websites from a single prompt, produce DESIGN.md documentation, enhance vague UI prompts, convert designs to React/shadcn-ui components, and generate walkthrough videos via Remotion. Use when the user needs AI-assisted UI design generation, prompt refinement, or screen-to-code workflows. Triggers on: stitch, stitch-design, stitch-loop, enhance-prompt, react-components, remotion, shadcn-ui, screen generation, ui generation.

Capture or repair reliable full-page screenshots for lazy-loaded, scroll-animated, Framer, WebGL/canvas, or reveal-heavy web pages. Use when full-page screenshots are blank, gray, white, sparse, show a tiny content strip, disagree with a working scroll video, or when article evidence/section crops must be derived from a trustworthy full-page image.

Install, configure, and operate Strix for AI-driven application security testing. Use when you need to run authorized vulnerability scans against local codebases, GitHub repositories, staging URLs, domains, or CI pipelines; configure Docker and LLM providers; choose quick, standard, or deep scan depth; or pass authenticated testing instructions to Strix. Triggers on: strix, ai pentest, vulnerability scan cli, appsec scan, bug bounty automation, strix ci, strix docker, strix scan mode, strix instruction file, headless security scan.

Summarize a customer interview transcript into a structured template with JTBD, satisfaction signals, and action items. Use when processing interview recordings or transcripts, synthesizing discovery interviews, or creating interview summaries.

Summarize a meeting transcript into structured notes with date, participants, topic, key decisions, summary points, and action items. Use when processing meeting recordings, creating meeting notes, writing meeting minutes, or recapping discussions.

Install and use Supabase Agent Skills (`supabase/agent-skills`) with AI coding agents. Covers install modes, skill selection, plugin path, verification, and safe fallback for direct Supabase CLI/database workflows.

Run a bounded cross-platform landscape scan before planning or implementation. Use when the real job is researching what exists, how people work around it, which solutions repeat, or how platform/tooling patterns map before deciding what to build. Produce reusable `.survey/{slug}/` artifacts, validate the artifact contract, and route planning or execution outward only after the survey is done.

Perform a detailed SWOT analysis — strengths, weaknesses, opportunities, and threats with actionable recommendations. Use when doing strategic assessment, competitive analysis, or evaluating a product or business position.

Build reproducible developer environments for real projects: local toolchains, runtime versions, Docker Compose or dev containers, onboarding flows, local service parity, bootstrap scripts, and environment troubleshooting. Use when the user needs a repo to run consistently across machines, containers, or staged environments, even if they only say setup dev environment, onboarding, Docker, devcontainer, local parity, bootstrap, or make this runnable. This is the canonical broader environment-setup skill. Route narrower `.env` and app-config questions to `environment-setup`.

Turn fuzzy work into one honest estimate packet by choosing the right sizing horizon, naming confidence and uncertainty, and separating discovery from delivery before anyone treats the number like a promise. Use when the user needs story points, t-shirt sizing, planning-poker prep, forecast-safe language, split-or-spike guidance, or cross-functional sizing across developer workflow, web/fullstack, product/ops, marketing/GTM, or game work. Route decomposition to `task-planning`, daily coordination to `standup-meeting`, and retrospective process learning to `sprint-retrospective`.

Turn vague features, bug clusters, roadmap items, launch work, or playtest findings into one execution-ready planning packet by choosing the right packet type, separating discovery from delivery, and making blockers, dependencies, and the next move explicit. Use when the user needs backlog cleanup, feature slicing, sprint or milestone prep, release planning, or roadmap-to-delivery translation across developer workflow, web/fullstack, product/ops, marketing/GTM, or game work. Also owns the retired `sprint-plan` name: sprint capacity from velocity and availability, Definition-of-Ready story selection, dependency and critical-path mapping, sprint goal, sprint plan summary. Route sizing to `task-estimation`, issue-state governance to `triage`, plan review to `plannotator`, daily syncs to `standup-meeting`, retros to `sprint-retrospective`, and pre-planning concept framing to `bmad`, `bmad-idea`, or `bmad-gds`.

Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.

Teach the user a new skill or concept, within this workspace. Use when the user wants to learn a concept or skill inside this workspace. Triggers on: teach me, help me learn, explain this properly.

Write internal technical documentation for engineers and operators: technical specs, architecture docs, ADRs, runbooks, migration plans, and developer-facing implementation guides. Use when the main job is capturing a technical decision, system boundary, operating procedure, or rollout path for builders and maintainers. Triggers on: tech spec, design doc, architecture doc, ADR, runbook, migration guide, implementation guide, rollout doc, operational guide, and internal technical writing. End-user onboarding guides, tutorials, FAQs, and help-center flows belong here too. Route API portals to `api-documentation`, release notes to `changelog-maintenance`, decks to `presentation-builder`, and GTM messaging to `marketing-automation`.

Create comprehensive test scenarios from user stories with test objectives, starting conditions, user roles, step-by-step actions, and expected outcomes. Use when writing QA test cases, creating test plans, defining acceptance tests, or preparing for feature validation.

Turn test-policy ambiguity into one packet-first validation brief. Use when the main job is deciding which gate is actually being shaped (merge, release, or scheduled), what evidence a change needs, how flaky or expensive suites should be handled, and whether the next owner is `backend-testing`, `debugging`, `code-review`, `deployment-automation`, `steam-store-launch-ops`, `game-ci-cd-pipeline`, `web-accessibility`, or `performance-optimization` instead of absorbing all test work here. Triggers on: test strategy, merge gate, required status checks, release gate, flaky-suite policy, regression policy, validation brief, release confidence, and what should we test.

Implement and debug Three.js motion with AnimationMixer, clips and actions, GLTF animation playback, skeletal rigs, morph targets, cross-fades, and frame-rate-independent procedural animation. Use when playing or blending model animations, driving bones or morphs, creating timeline motion, or fixing animation lifecycle and performance issues. Triggers on: Three.js animation, AnimationMixer, AnimationAction, AnimationClip, GLTF animation, skeletal animation, bones, morph targets, crossfade, procedural motion.

Build and debug the Three.js scene foundation: renderer setup, cameras, scene graph, transforms, resize handling, color management, render loops, and resource disposal. Use when creating a Three.js scene, choosing a camera or renderer, fixing coordinate space or hierarchy bugs, or establishing a reliable WebGL canvas baseline. Triggers on: three.js scene, WebGLRenderer, PerspectiveCamera, OrthographicCamera, Object3D, scene graph, camera setup, resize canvas, render loop, coordinate system, transform.

Build and optimize Three.js geometry with built-in primitives, BufferGeometry, BufferAttributes, indexed meshes, custom vertex data, lines and points, and instanced rendering. Use when creating meshes, editing vertices or UVs, generating shapes, reducing draw calls, or diagnosing geometry memory and culling behavior. Triggers on: Three.js geometry, BufferGeometry, BufferAttribute, vertices, indices, custom mesh, InstancedMesh, instancing, draw calls, line geometry, points, mesh optimization.

Implement and debug Three.js interaction with Raycaster, pointer and touch coordinate conversion, hover/selection state, camera controls, dragging, keyboard input, and world-to-screen projection. Use when picking objects, adding OrbitControls or PointerLockControls, handling mouse/touch input, or fixing interactive 3D behavior. Triggers on: Three.js interaction, Raycaster, raycast, picking, click mesh, hover, object selection, OrbitControls, drag controls, pointer events, touch controls.

Design and debug Three.js lighting with directional, point, spot, hemisphere, and area lights, shadow maps, image-based lighting, environment maps, light helpers, and performance budgets. Use when lighting a 3D scene, configuring shadows, setting up HDR illumination, matching a visual reference, or fixing dark, flat, or expensive renders. Triggers on: Three.js lighting, directional light, point light, spot light, shadows, shadow map, ambient light, HDR environment, IBL, environment map, light helper.

Load and manage Three.js assets with LoadingManager, GLTFLoader, DRACOLoader, KTX2Loader, texture/HDR loaders, async error handling, caching, progress reporting, and resource ownership. Use when importing GLB/GLTF models, textures, HDR assets, or compressed geometry, or when fixing slow, failed, duplicated, or leaky asset loads. Triggers on: Three.js loader, GLTFLoader, GLB, glTF, DRACOLoader, KTX2Loader, LoadingManager, texture loader, HDR loader, loading progress, asset cache.

Choose, configure, and optimize Three.js mesh materials: basic, Lambert, Phong, Standard, Physical, toon, points, lines, and shader-backed surfaces; PBR maps, transparency, environment reflections, cloning, and disposal. Use when styling meshes, tuning PBR, fixing transparency or material sharing bugs, or reducing material cost. Triggers on: Three.js material, MeshStandardMaterial, MeshPhysicalMaterial, PBR, roughness, metalness, transparency, environment map, mesh surface, material clone.

Build and optimize Three.js post-processing with EffectComposer, RenderPass, bloom, anti-aliasing, SSAO, depth of field, outlines, color correction, custom ShaderPasses, render targets, resize handling, and frame-budget controls. Use when adding or fixing screen-space effects, composer pass order, render target sizing, or visual-effect cost. Triggers on: Three.js postprocessing, EffectComposer, RenderPass, UnrealBloomPass, FXAA, SMAA, SSAO, depth of field, bloom, ShaderPass, render target, screen-space effect.

Create and debug Three.js shaders with ShaderMaterial, RawShaderMaterial, uniforms, varyings, GLSL vertex and fragment programs, texture sampling, procedural effects, onBeforeCompile extensions, shader chunks, instancing, and GPU performance checks. Use when writing custom GLSL, extending a built-in material, animating shader uniforms, or diagnosing shader compile, coordinate-space, and rendering issues. Triggers on: Three.js shader, ShaderMaterial, RawShaderMaterial, GLSL, uniforms, varyings, onBeforeCompile, fragment shader, vertex shader, fresnel, displacement, shader chunk.

Load, configure, and optimize Three.js textures: color space, wrapping, filtering, mipmaps, UV transforms, normal/AO/PBR maps, cube and HDR environments, data/canvas/video textures, render targets, texture memory, and disposal. Use when images look wrong, UVs repeat incorrectly, HDR/IBL is needed, or texture quality and GPU memory need tuning. Triggers on: Three.js texture, TextureLoader, UV mapping, color space, sRGB, mipmap, anisotropy, normal map, environment map, cubemap, HDR texture, render target, video texture.

Turn a decision you can't fully answer into a questionnaire for someone else to fill in. Use when a decision needs another person's input before it can be settled. Triggers on: to- questionnaire, make this a questionnaire, ask them these questions.

Turn the current conversation into a spec and publish it to the project issue tracker: no interview, just synthesis of what you've already discussed. Use when the conversation is ready to become a written spec. Triggers on: to-spec, to-prd, write this up as a spec.

Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker (edges as text in one file per ticket locally, or native blocking links on a real tracker). Use when a plan or spec should become executable work items. Triggers on: to-tickets, to-issues, to-plan, break this into tickets.

Set up, run, and contribute to TokHub (github.com/yaojingang/TokHub) — an open-source AI API relay monitoring, recommendation, and OpenAI-compatible gateway system with L1/L2/L3 channel health probing, usage metering, alerts, audit, and Docker self-hosting. Use when the user asks about TokHub, "AI API 中转站监控", cloning/running the Go + React monorepo (TOKHUB_ROLE, sqlc, TimescaleDB, NATS), the L1/L2/L3 probe algorithm, the OpenAI-compatible `/gateway/v1/*` endpoint, or contributing a PR to TokHub. Do not use for connecting a running agent to a live TokHub instance's own API (that is covered by the project's own bundled `agent-skills/tokhub` skill inside the TokHub repo, not this one).

Stand up a self-hostable, typo-tolerant search environment with Typesense — the open-source Algolia / ElasticSearch alternative (single C++ binary, <50ms instant search, no runtime deps). One routing-first skill: pick a server mode (binary download, official Docker image, or managed Typesense Cloud), install an API client (Python/JS/PHP/Ruby official; Go/Dart/C# community), design a collection schema, index documents, and run searches with typo tolerance, faceting/filtering, geo-search, sorting, grouping, synonyms, curation, scoped API keys, and federated multi-search — then wire an InstantSearch.js UI and a Raft-based HA cluster for production. Use when the user wants to build or operate an installable search backend, add site/app/product search, or migrate off Algolia/Elasticsearch. Triggers on: typesense, search engine, typo-tolerant search, algolia alternative, elasticsearch alternative, instantsearch, faceted search, geo search, vector search, self-hosted search, site search, product search.

Inspect Underworld Overseer JSON saves and generate or troubleshoot interactive dungeon maps with RobThePCGuy/Underworld-Overseer-Save-Mapper. Use when a user wants to validate a save's Map records, visualize room coordinates and DescriptorID values, install the mapper, or diagnose empty/broken HTML output. Triggers on: Underworld Overseer save, dungeon save map, DescriptorID, Map JSON, Underworld Overseer mapper, or save-to-HTML.

Automatically rig 3D models with UniRig (VAST-AI-Research, SIGGRAPH'25) — predict a skeleton, predict skinning weights, and merge the rig back onto the original mesh. Use when the user wants auto-rigging for .obj/.fbx/.glb/.gltf/.dae/.vrm assets, a skeleton or skin weights for a character or creature, a UniRig environment prepared on a CUDA machine, batch rigging of a model directory, or an honest comparison between UniRig, SkinTokens, Tripo, Mixamo, AccuRig, and Blender Rigify. Triggers on: unirig, auto rig, auto-rigging, 3D rigging, skeleton prediction, skinning weights, rig a character, armature generation, rigged glb, rigged fbx, bone weights.

Install, configure, and use the Unity Command Line Interface (CLI) for automated production workflows, project management, and cloud integration. Use when setting up CLI-based build automation, CI/CD pipelines, Editor/module management, authentication, or localhost API calls via the experimental Unity Pipeline package. Designed for verifiable, machine-readable game production workflows where the build machine must be describable and tests must return evidence. Triggers on: Unity CLI, unity command line, unity automation, unity ci/cd, unity build script, unity pipeline, unity production workflows.

Evaluate and adopt Unity game-development skill packs from external repositories into a safe, reusable local package. Use when maintainers want to import game-dev workflows (Addressables, Cinemachine, GAS, VContainer, UniTask, Wwise, etc.) without blindly trusting third-party prompts.

Audit, selectively install, route, and maintain the official Unity-Technologies/skills collection for Unity Editor, CLI, UI, rendering, physics, localization, multiplayer, UGS, IAP, and LevelPlay workflows. Use when the user explicitly names Unity-Technologies/skills, Unity Skills, or the official Unity agent-skill pack; wants its real inventory; needs a pinned selective install or refresh; must choose the correct upstream sub-skill; or needs to validate the pack before adoption. Inspect actual skill directories, support files, frontmatter, license, and destination collisions before any copy. Require separate approval before Editor or module installation, live C# evaluation, project mutation, cloud deployment, billing, ads, source-control publication, or credential use. Route generic third-party Unity pack curation to `unity-gamedev-skill-pack`, CLI-only work to `unity-cli`, build failures to `game-build-log-triage`, and game CI design to `game-ci-cd-pipeline`.

Use when you need to pick high-quality Unsplash images for product/design assets (avatars, headshots, portraits, large website backgrounds, and abstract wallpapers) and output real Unsplash URLs plus practical instructions for producing the right resolutions and aspect ratios (1:1, 4:5, 3:4, 16:9, 9:16).

Turn a weak/cheap "Flash" model into a "Pro" performer by wrapping HKUDS UpSkill — captures agent session failures, has a strong Teacher model analyze them and draft a skill, then validates it against the weak Student model in a closed Ralph Loop (up to 3 rounds) before storing it for automatic reuse. Use when the user wants to install UpSkill, run `/upskill-init`, `/upskill-configure`, `/upskill-build`, `/upskill-run`, `/upskill-list`, `/upskill-status`, `/upskill-mode`, `/upskill-model`, `/upskill-remove`, or `/upskill-uninstall`, wants a cheap model to perform closer to a Pro model without switching, or wants a good session (success or failure) distilled into a validated skill. Triggers on: upskill, up-skill, flash to pro, teacher student distillation, ralph loop skill validation, distill agent failures into skills. Routes skill-quality ratcheting to `skill-autoresearch`, scaffolding to `write-a-skill`, and spec-compliance rewrites to `skill-standardization`.

Create refined user personas from research data — 3 personas with JTBD, pains, gains, and unexpected insights. Use when building personas from survey data, creating user profiles from research, or segmenting users for product decisions.

Segment users from feedback data based on behavior, JTBD, and needs. Identifies at least 3 distinct user segments. Use when segmenting a user base, analyzing diverse user feedback, or building a segmentation model.

Create user stories following the 3 C's (Card, Conversation, Confirmation) and INVEST criteria with descriptions, design links, and acceptance criteria. Use when writing user stories, breaking down features into backlog items, or defining acceptance criteria.

Generate value proposition statements for marketing, sales, and onboarding from existing value propositions. Use when writing marketing copy, creating sales messaging, or crafting onboarding messages.

Design a detailed value proposition using a 6-part JTBD template — Who, Why, What before, How, What after, Alternatives. Use when creating a value proposition, analyzing customer value delivery, or articulating why customers should choose your product.

Run Vercel-specific deployment operations for linked web/fullstack projects: preview deploys, direct production deploys, staged deploy + promote flows, aliases/domains, environment-variable sync, and rollback response. Use when the request is specifically about operating a project on Vercel after the provider is already chosen — especially when someone needs a preview URL, stable alias, custom domain, env-scope fix, production cutover, or rollback on Vercel. Triggers on: Vercel deploy, Vercel preview URL, vercel promote, vercel alias, Vercel domain, Vercel env, Vercel rollback, and Vercel CLI deployment. Route provider-neutral release strategy to `deployment-automation`, CI/workflow authoring to `deployment-automation`, and local install/auth/bootstrap work to `system-environment-setup`.

Plan and route programmable or automated video production across code-first, template-first, and hybrid content pipelines. Use when the user needs repeatable video generation, branded short-form content, personalized videos, social clip batches, captioned/localized variants, video APIs, or video creation from data and templates — even if they only say video production. Also the direct owner of explicit Remotion requests (React video compositions, scenes, render workers) and of the retired `remotion-video-production` name. Triggers on: Remotion, Remotion render pipeline, React video composition, programmatic video, automated video creation, video API, personalized video, batch-create shorts, render videos from code, repurpose content into clips.

用镜头配方卡 + 已验收模板 + 代码/音频资产制作电影感产品视频（Remotion + 真实页面截图 + 2.5D 运镜 + 节奏卡点 + 声音设计）。 Use when making a cinematic product promo or desktop demo video, naming the Ink Press template, reproducing template video effects, or building single motion-design shots with recipe cards. Triggers on: video-shotcraft, Ink Press template, product video, cinematic promo, shot recipe card, remotion video.

Turn a reference video into a super detailed recreation or inspiration prompt. Use when the user provides, mentions, uploads, links, or points to a video and asks to analyze the design, UI, animations, transitions, scroll interactions, typography, colors, assets, WebGL/Three.js, storytelling, section-by-section behavior, or to create a prompt/article that recreates the page, app, interaction, or motion system.

Turn ONE topic into a finished Vox-style paper-collage explainer / ad video, end to end on the Atlas Cloud API + local ffmpeg — script, collage keyframes, motion, voice-over, music, captions, all automated. Use this whenever the user wants a "Vox style" video, a paper/torn-paper collage animation, a "motion collage", a narrated explainer or short ad built from AI-generated collage posters, a scrapbook-style tribute, or wants to turn a topic / product / person into a punchy narrated collage video — even if they don't say the word "Vox". Also use when reproducing Stav Zilber / rom1trs / Higgsfield-style collage ad workflows, or when the user asks for a motion collage or a scrapbook-style tribute. Triggers: "vox video", "collage video", "motion collage", "paper collage explainer", "make a collage ad", "turn this topic into a collage video".

Route web-game auto-playtesting with WAI Play (waiterve/wai-play): decide whether the next move is a testability check, authoring or repairing the `GameFlowAgentAPI` bridge, running a real browser playtest, reading the five-dimension quality report, or unblocking a key node the agent cannot reach. Use when the user wants an AI agent to actually play their HTML5 / canvas / vibe-coded web game and return reproducible evidence, scores, and fix suggestions across the five supported types (survivor-like, arcade shooter, platformer, puzzle/card, visual novel). Triggers on: wai-play, WAI Play, auto-playtest, AI plays my game, web game testing agent, GameFlowAgentAPI, GameFlowIntegration, jumpToScenario, game quality score, playtest evidence. Route Unity/Unreal frame-time work to `game-performance-profiler`, engine build failures to `game-build-log-triage`, human playtest notes to `game-demo-feedback-triage`, and generic browser automation to `browser-harness`.

Stop. That last message did not land: re-pitch it. Use when the previous answer did not land and needs re-pitching in plain language. Triggers on: wait what, I don't follow, re-pitch that.

Strip multi-vendor AI provenance marks from text and files using the guillaumemeyer/watermarks-remover Python toolkit: invisible Unicode/space homoglyphs (Layer A, deterministic), statistical token-sampling watermarks via agent-guided rewrite (Layer B, best-effort), and C2PA/EXIF/XMP/doc-props metadata on PNG, JPEG, SVG, PDF, DOCX, ODT, HTML, and Markdown. Covers Claude, Gemini/SynthID-Text, OpenAI provenance surfaces, and open-LLM Kirchenbauer-style marks, plus optional external backends for SynthID pixel scoring and CtrlRegen pixel-domain removal. Use when the user wants to strip AI watermarks, remove C2PA / Content Credentials, clean AI metadata from a file, remove invisible Unicode / zero-width characters from AI-generated text, or audit a directory/website for AI provenance signals. Triggers on: "remove watermark", "strip C2PA", "remove AI metadata", "clean invisible unicode", "remove-ai-marks", "SynthID removal", "audit_dir.py", "Layer A / Layer B watermark removal".

Plan a huge chunk of work (more than one agent session can hold) as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear. Use when the work is too large for one agent session and needs a shared decision map first. Triggers on: wayfinder, map this out, too big to plan in one go.

Routing-first skill for web accessibility remediation and verification. Use when the main job is classifying which accessibility surface is failing — semantics, keyboard/focus, labels/announcements, visual perception/reflow, media alternatives, or routed-app navigation feedback — and turning vague audit, WCAG, axe/Lighthouse, or "make this accessible" requests into one concrete remediation packet. Broad UI critique for hierarchy, polish, and launch readiness belongs here too. Route component API architecture and system governance to `design-system`, and responsive layout strategy to `responsive-design`.

Route web design, motion, WebGL, and visual-styling requests into the narrowest matching sub-skill of the 79-skill MengTo/Skills web-design family — Style/Design-System Packs, Full-Site Quality & Direction, Motion & Scroll Systems, WebGL/3D Backgrounds & Shaders, Library Embeds & Integrations, Component & State Effects, and CSS Technique Primitives. Use when picking and fetching the right visual recipe for a marketing site, landing page, portfolio, or dashboard shell, then combining layers in direction → layout → background → motion → component-state → polish order. Triggers on: web design, landing page, awwwards, GSAP, Lenis, scroll animation, WebGL background, shader, glass UI, dark mode design system, editorial layout, cursor trail, progressive blur, gradient border, three.js background, vanta, globe.gl, unicorn studio, marketing site polish.

Route Three.js and browser game-development requests into the narrowest matching sub-skill of the 19-skill MengTo/Skills game-development family — level authoring, map editing, cameras, enemy systems and AI, monster rigs, action combat, encounter design, inventory, hybrid/Vesperfall assets, VFX, audio feedback, mobile controls, performance tuning, QA, changelogs, and release. Use when a task touches any of those for a playable web game, and fetch the matching sub-skill before doing the work instead of improvising. Route Unity/C# engine work to `unity-gamedev-skill-pack` instead. Triggers on: three.js game, browser game, webgl game, isometric arpg, enemy AI, game camera, game inventory, game VFX, game audio feedback, ship web game, playtest web game, mobile threejs game, game encounter design, monster rig, hybrid game assets.

Run an end-to-end webtoon production harness — a 27-agent / 4-phase Claude Code agent team that takes a single episode from trend research to a finished vertical-scroll viewer. Use when the user wants to produce a webtoon episode: research popular webtoon trends, write a dialogue-heavy, high-tension, twist-every-episode scenario, render character reference sheets first, render 50+ panels per episode with in-image speech-bubble (Korean dialogue) baking via codex-image, run a generation–validation loop, and assemble a vertical-scroll viewer. Triggers on: webtoon harness, make a webtoon, produce a webtoon episode, webtoon scenario to image, next episode, redraw panel, stronger twist, webtoon trend research. Web data gathering for the trend-research phase routes through the `scrapling` skill. Pure webtoon recommendation/critique is answered directly.

Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.

Creates reusable SKILL.md-based agent skills through requirements, drafting, and user review; use to formalize a workflow or add a new agent capability.

Write, rewrite, review, or continuously refine X/Twitter posts in Meng To's current voice using his deduplicated authored-post corpus, personal and product context, shared resources, and Content repo evidence. Use when asked to improve a tweet in Meng's tone, draft an original post, reply, thread, resource share, or quote post as Meng, check whether an angle repeats an earlier post, study Meng's writing style, ingest another 20-50 authored posts, or hone the reusable Meng X voice profile.

Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.

Create product backlog items in Why-What-Acceptance format — independent, valuable, testable items with strategic context. Use when writing structured backlog items, breaking features into work items, or using the WWA format.

Check a user's latest X/Twitter bookmarks and turn recent saved posts into source-backed quote-post drafts calibrated against the user's latest 100 authored posts. Use when asked to review X bookmarks, create quote posts from bookmarks, refresh a bookmark quote queue, run a bookmark quote automation, study a user's X voice, or write first-person quote posts from X sources.

Use when the user needs X (Twitter) data or confirmation-gated X actions through Xquik: tweet search, user lookup, follower extraction, media download, monitoring, webhooks, MCP, SDKs, posting, likes, DMs, and profile updates. Requires a Xquik API key. Never ask for X login material.

Turn YouTube growth advice videos into an evidence-checked, executable 90-day channel operating plan. Use when the user asks how to grow a small YouTube channel, how to survive the early no-views phase, how to design titles and thumbnails, whether an algorithm claim is true, or how to run packaging experiments with YouTube Studio only. Also triggers on: 떡상, 유튜브 성장, 죽음의 계곡, 알고리즘 미신, 채널 일관성, 제목 썸네일 A/B 테스트.

国内自媒体发布前风险自审与保意修复：审口播稿、文章、图文笔记、字幕、封面文字能不能发抖音/小红书/微信视频号，给出具体位置、依据和可直接替换的改稿；被限流/删除/处罚后帮你复盘归因；你的行业敏感词、误报白名单和踩坑案例会沉淀成个人规则库，越用越准。Use when 用户说"能不能发""审一下稿子""查违禁词/敏感词""会不会限流/被限流了""帮我改成能发的版本""发布前检查""被平台处罚/删除了""帮我盯着这个词""导入违禁词表"。Not for 海外平台(X/YouTube)内容审核、起号涨粉策略、写稿创作本身。

Operate the-open-engine/zeroshot, the MIT-licensed agent orchestration system that routes a software task through a persistent graph of provider CLI agents, validators, worktree or Docker isolation, and optional PR delivery. Choose the established Node `zeroshot` product, the standalone `zeroshot-rust` engine, or its typed Python SDK; audit provider and credential boundaries; validate configs without paid calls; freeze acceptance, isolation, delivery, and cost before execution; monitor or recover durable runs; and verify trace or semantic JSONL evidence without exposing content. Use when the user names ZeroShot, zeroshot-rust, The Open Engine, conductor-bootstrap, `zeroshot run`, multi-agent coding workflow, ZeroShot trace export, or a ZeroShot target. Never start, resume, schedule, force-stop, create a PR, ship, clean, purge, or update without the matching approval.

Enterprise AI adoption platform for Claude Code — measures team usage via OpenTelemetry, syncs skills/MCP servers/hooks from a central dashboard, and delivers context-aware skill suggestions at prompt time. Solves the Intention-Action Gap: organizations that deploy Zeude see 3x adoption improvement (6% → 18%). Requires Supabase (config) and ClickHouse (analytics). Triggers on: zeude, ai adoption, claude code adoption, enterprise claude, opentelemetry claude, skill sync, team claude management.

Get higher-level architectural perspective: maps modules, callers, dependencies using domain vocabulary

Drive zvec-grep (`zg`), zvec-ai's local-first workspace search CLI and MCP server, to index code and documents, run hybrid, vector, BM25, or exhaustive managed ripgrep searches, configure agent integrations, and diagnose index or server freshness. Use when a task needs local workspace-grounded retrieval across code, Markdown, text, structured data, or images. Triggers on: zvec-grep, zvec grep, `zg`, @zvec/zvec-grep, local-first workspace search, hybrid workspace search, semantic repository search, zvec MCP, zvec_grep_search, managed ripgrep, `.zvec-grep`, or Remote Embedding authorization.