The library

portainer-mcp-hygiene · Large Portainer queries where 20K+ tokens of unneeded data would waste context.

eval-harness · AI-assisted workflows where deterministic pass/fail is needed before implementation.

evals · Teams validating agent prompt changes or model upgrades with quantified reliability metrics.

evm-token-decimals · DeFi parsing tasks where token amounts must be normalized by contract decimals.

k8s-capi · Use when working with skill named k8s-capi.

exa-search · Use when working with skill named exa-search.

k8s-certs · Use when working with skill named k8s-certs.

expo-api-routes · Use when working with skill named expo-api-routes.

k8s-cilium · Troubleshooting cross-pod connectivity issues faster than raw iptables inspection

expo-cicd-workflows · Ensuring EAS workflows conform to latest schema without manual spec parsing

release-skills · Packaging multi-language projects (Node/Python/Rust/Claude Plugin) with unified versioning

k8s-cli · Exploring available kubectl-mcp tools and testing CLI invocation without code

expo-core · Aligning Expo app architecture decisions across team and avoiding mixed SDK versions

k8s-core · One-shot inspection and management of pods, configmaps, secrets without kubectl CLI

expo-deployment · Automating end-to-end iOS/Android store submission without manual Xcode/Android Studio

expo-dev-client · Adding custom native code without ejecting to bare workflow

k8s-deploy · Declarative deployment without kubectl CLI or git-ops operator overhead

expo-module · Creating reusable cross-platform native extensions within Expo ecosystem

k8s-diagnostics · Root-cause analysis of pod failures faster than manual log + event review

expo-orchestrator · Reducing multi-platform release ceremony with parallel builds and status aggregation

k8s-gitops · Maintaining cluster state in version control without manual kubectl apply

expo-tailwind-setup · Unifying web and native styling via Tailwind without parallel CSS systems

extra-languages-core · Enforcing native-idiom-first + contained-boundary rule across niche-stack specialists

k8s-incident · Fast triage of production Kubernetes failures (pod, node, network, storage, control-plane)

extra-languages-orchestrator · Deciding if a platform-specific task belongs in niche-stack cluster and delegating to correct specialist

baoyu-electron-extract · Inspecting shipped Electron desktop app structure, dependencies, and implementation details

k8s-kind · Local development and CI/CD testing without cloud infrastructure, multi-node setups

k8s-kubevirt · Specialized use case detailed in body

k8s-multicluster · Specialized use case detailed in body

fastapi-patterns · Specialized use case detailed in body

k8s-networking · Specialized use case detailed in body

k8s-operations · Specialized use case detailed in body

file-uploads · Specialized use case detailed in body

k8s-policy · Enforcing cluster-wide security rules (no privileged containers, require labels) without human review on every deployment.

k8s-rollouts · Rolling out new versions to a subset of traffic first (5% → 10% → 20%...) while metrics verify no regression.

firebase · Building a backend for a React/Vue app when you want to skip the server and denormalize data for queries.

k8s-security · Finding which service accounts have admin access and enforcing no-wildcards policy before production.

k8s-service-mesh · Shifting traffic between service versions (A/B testing) or enabling mutual TLS without modifying application code.

k8s-storage · Provisioning persistent storage for databases or file shares when you need volumes that survive pod restarts.

flox-environments · Setting up project dependencies (Python, Node, Rust, system libs) once and sharing a lock file so all devs get identical environments.

k8s-troubleshoot · Quickly diagnosing why a pod won't start or a container keeps crashing without ssh-ing into nodes.

flutter-dart-code-review · Auditing Flutter code for state management anti-patterns and widget best practices regardless of which state library is used.

k8s-vind · Creating lightweight isolated Kubernetes clusters for CI/CD, testing, or multi-tenant development without spinning up full clusters.

foundation-models-on-device · building privacy-preserving or offline AI features, extracting structured data, or streaming model output in a native iOS app.

ai-model-nodejs · Only SDK supporting image generation (ai.createImageModel + generateImage). Text models via ai.createModel with groups cloudbase, hunyuan-exp, or custom-*. Model IDs (deepseek-v4-flash, deepseek-v3.2,

fp-ts-errors · you want type-safe explicit error handling in TS APIs or services.

ai-model-web · Default routing for page/页面/Web/前端/frontend/网页/H5 AI — call directly from browser, do NOT propose a Node.js proxy. Covers generateText and streamText. Models via ai.createModel with groups cloudbase,

fp-ts-pragmatic · writing or refactoring TypeScript with fp-ts (Option/Either/TaskEither).

ai-model-wechat · Features generateText and streamText with callbacks (onText, onEvent, onFinish). Models via wx.cloud.extend.AI.createModel with groups hunyuan-exp (小程序成长计划), cloudbase (main managed), or custom-*. Mod

fp-ts-react · building React apps with functional, type-safe state and fetch handling.