opsskills

Plugin opsskills from marketplace claude-development-skills in VelimirMueller/claude-development-skills. Commands read from the repo's manifest — 23 days ago.

Skills (80)

• build-llm-seam

  Use when adding the first LLM call to a service, or when model calls are scattered through it — one module owns every call - model tiers from config, retries, validated structured output, streaming, prompt caching, OTel gen_ai spans with cost, versioned prompt files, PII redaction.

• build-mcp-server

  Use when exposing a service or dataset to AI clients through the Model Context Protocol — few task-shaped tools with typed inputs, read-only by default, server-enforced confirmation for writes, stdio or Streamable HTTP, OAuth for remote, size limits, tests with the Inspector.

• build-rag-pipeline

  Use when building retrieval-augmented answers over your own documents — structure-aware chunking, pinned embeddings in pgvector, hybrid vector plus full-text search with RRF, tenant isolation in the query, citations, an "I don't know" path, retrieval evals before prompt tuning.

• secure-llm-features

  Use when shipping or reviewing an LLM feature that reads untrusted text or has tools — prompt-injection threat model, least-privilege tools scoped by the user's token, safe output handling, per-user rate and cost limits, logs without prompt PII; mapped to the OWASP LLM Top 10 (2026).

• set-up-llm-evals

  Use when an LLM feature ships on vibes — golden datasets in the repo, deterministic checks first, an LLM judge with a rubric calibrated against human labels, a small CI regression gate, a cost budget and a results history that survives a model or prompt change.

• build-nextjs-backend

  Use when building or reviewing the server side of a Next.js 16 App Router app — a server-only data access layer that authorizes every call, Server Actions as public endpoints, route handlers, caching with Cache Components, proxy.ts, env validation and error boundaries.

• build-supabase-edge-function

  Use when logic must run server-side next to Supabase — a webhook, a cron or database-triggered job, a third-party call with secrets, or an API for non-Next clients — and you must choose and build an Edge Function with caller auth, CORS, secrets and tests.

• configure-backend-tests

  Use when a backend service needs a test setup or its tests mock the database. Builds the pyramid - pure unit tests, real-Postgres integration via testcontainers, HTTP-level tests, OpenAPI contract tests, factories and CI-ready layout (Vitest, go test, pytest).

• design-http-api

  Use when adding or changing HTTP endpoints, or when the frontend needs typed calls — sets the OpenAPI 3.1 contract, RFC 9457 errors, cursor pagination, idempotency keys, versioning, and a generated typed client behind the frontend fetcher seam.

• harden-backend

  Use when a backend service is about to take untrusted traffic or fails a security review. Adds edge validation, rate and size/time limits, a CORS allow-list, API security headers, SSRF-safe outbound fetch, safe error responses and CI scanning (Hono, Go, FastAPI).

• scaffold-fastapi-service

  Use when starting a Python HTTP API or adding a service to a repo with no FastAPI app yet — scaffolds a uv src-layout FastAPI service with lifespan, pydantic-settings, ruff, mypy, pytest, health/readiness, problem+json errors and an async SQLAlchemy session dependency.

• scaffold-go-service

  Use when starting a Go HTTP service or adding one to a repo with no Go module yet — scaffolds stdlib net/http (1.22+ routing), cmd/ and internal/ layers, slog JSON logs, validated env config, problem+json errors, health/readiness, graceful shutdown and a golangci-lint v2 config.

• scaffold-hono-service

  Use when starting a TypeScript HTTP API or adding a service to a repo with no Hono app yet — scaffolds a strict-TS Hono service with validated config, problem+json errors, OpenAPI from route schemas, health/readiness and graceful shutdown, runnable on Node.

• secure-supabase-rls

  Use when adding or changing a Supabase table, view, function or storage bucket exposed through the Data API, or auditing one — enables RLS, writes per-operation role-scoped policies, indexes them, keeps definer functions private, and proves allow and deny with pgTAP.

• set-up-backend-auth

  Use when a backend service must authenticate callers or enforce access. Verifies OIDC JWTs against JWKS, adds BFF session cookies for browsers, puts authorization in service-layer policies, denies by default and scopes every query to the tenant (Hono, Go, FastAPI).

• set-up-background-jobs

  Use when a backend must do work outside the request (email, webhooks, imports, schedules) or publish events reliably — adds a Postgres-backed queue (pg-boss, River, procrastinate), the transactional outbox, idempotent handlers, retries with dead letter, cron jobs.

• set-up-database

  Use when a backend needs Postgres, or its schema changes by hand or by `db push` — sets up local Postgres via compose, the ORM or sqlc, pooling, service-layer transactions, versioned reviewed migrations, seed data, and a migration-safety gate.

• set-up-nextjs-supabase-auth

  Use when adding or fixing Supabase Auth in a Next.js 16 App Router app — @supabase/ssr cookie clients, session refresh in proxy.ts, verified getClaims() on the server, protected routes, PKCE callback and email-link routes, and sign-in by password, magic link or OAuth.

• set-up-observability

  Use when a backend service should emit traces, metrics and trace-correlated logs. Wires app-side OpenTelemetry - SDK started before app code, HTTP and DB auto-instrumentation, use-case spans, RED metrics, OTEL_* env only, flush on shutdown (Hono, Go, FastAPI).

• set-up-supabase

  Use when a project starts using Supabase or its supabase/ folder is missing or half-set-up — wires the CLI local stack, migrations as the only schema path, generated types, per-environment config, and one client seam per context so the secret key never reaches a bundle.

• build-cli

  Use when creating or extending a command-line tool in TypeScript, Go, Python or Rust — subcommands, typed args, config precedence, stdout/stderr split, --json, exit codes, NO_COLOR/TTY handling, --dry-run, and spawn-the-binary tests.

• release-cli

  Use when cutting a release of a CLI — SemVer over the CLI surface, changelog-driven release notes, tag-triggered CI, per-language distribution (npm, GoReleaser, PyPI, cargo-dist), signed or provenance-bearing artifacts, and install plus verify instructions in the README.

• set-up-dev-toolchain

  Use when setting up or auditing a repo's developer toolchain — mise pins and tasks or just, shared task verbs, lefthook git hooks, editorconfig, GitHub CLI conventions, and a doctor task that verifies everything.

• audit-security

  Use when asked to review a repo's security, before a release or go-live, or after adding auth, uploads, or a new integration. Scans secrets, dependencies, authz, headers and CI supply chain against security-baseline.md; reports ranked findings, fixes only what is approved.

• audit-toolchain

  Use when asked whether a repo's tooling is current, before a toolchain migration, or when a dependency or build step feels dated. Compares the repo to tech-radar.md and the stack profile and reports gaps with cost; changes nothing unless asked.

• extend-skillset

  Use when the stack profile names a tool no installed skill covers, or the user asks for a new skill. Authors one in the house style, audit-first and profile-aware, into the project's .claude/skills or this plugin repo, then checks it.

• promote-release

  Use when deploying — promoting develop to stage (stg) or stage to main (prd) by fast-forward of a verified SHA, with the ancestry invariant, approval gates and deploy-status preconditions.

• set-up-stack-profile

  Use when starting in a repo with no .claude/stack-profile.md, when the stack changed, or when the user asks which plugins and skills to use. Detects the stack, asks only the gaps, writes the profile, and lists plugins and skill order.

• write-commit-messages

  Use when committing work or writing a commit message — produces a subject and body any developer from junior to CTO can read and act on.

• write-pull-requests

  Use when opening or updating a pull request — produces a bug-fix or feature description reviewers from junior to CTO can follow end to end.

• clean-frontend-scaffolding

  Use when a freshly scaffolded Vite app still has the demo App, default styles, sample assets and placeholder title; reduces them to a minimal shell before real structure is added.

• configure-accessibility

  Use when making a frontend accessible — turns on a11y linting (Biome a11y rules for React/JSX and, with full SFC support, for Vue templates), establishes semantic-HTML/focus/reduced-motion conventions, and adds axe checks to the test stack so accessibility regressions fail CI.

• configure-analytics

  Use when adding analytics to a frontend — one provider-agnostic seam (trackEvent/trackPageView), Core Web Vitals as real-user data, consent-gated by default (GDPR/TDDDG), path-only page views, no PII, no-op without config.

• configure-ci

  Use when setting up CI/CD for a frontend project — a least-privilege GitHub Actions pipeline (install, lint, typecheck, audit, test, build, e2e, bundle budget) that gates merges, plus preview deploys per pull request for the project's host (Netlify, Vercel, self-hosted).

• configure-error-tracking

  Use when wiring production error tracking - points the captureError seam at Sentry (tracing, masked replay, explicit data collection), uploads hidden source maps with the Vite plugin, tags release and environment; no-op without a DSN.

• configure-linting

  Use when setting up lint and format in a frontend project - installs Biome as the only linter, formatter, import sorter and Tailwind class sorter (replacing ESLint/Prettier), with Vue SFC support, a lefthook pre-commit hook and a CI check.

• configure-test-stack

  Use when setting up tests for a frontend project - Vitest (Node unit and integration, real-browser UI via Playwright), Storybook stories as tests, Playwright e2e and MSW, under tests/ or colocated per the stack profile.

• configure-typescript

  Use when setting up or hardening TypeScript in a frontend project - strict flags, noUncheckedIndexedAccess, TS 6/7-safe paths, and one @/ alias kept in sync across tsconfig, Vite, Vitest, Storybook and Playwright.

• create-module

  Use when adding new logic to a frontend project (helper, hook, composable, wrapper, store) - routes it to the right layer (utils, libs, hooks, stores, feature folder) so components stay thin, with a unit test.

• optimize-performance

  Use when tuning frontend performance — enables the React Compiler, splits code by route and lazy-loads heavy chunks, adds bundle analysis with a size budget, measures Core Web Vitals (LCP/INP/CLS) with web-vitals, and applies image/CLS best practices. Measure first, then optimize.

• scaffold-frontend-project

  Use when starting a new frontend app from an empty directory - scaffolds a Vite 8 + TypeScript SPA (React 19 or Vue 3.5), pins the Node LTS and package manager, and wires Tailwind v4, leaving a running base.

• set-up-auth

  Use when adding authentication to a frontend SPA — current user as TanStack Query server state, tokens never in localStorage (httpOnly cookies, or in-memory access + refresh cookie), login/logout mutations, one 401-refresh in the fetcher, route guards.

• set-up-design-system

  Use when establishing a design system — Tailwind v4 @theme tokens, class-based dark mode from a persisted theme store (CSP-safe pre-paint script), a cn() merger, and cva variant primitives; shadcn/ui is the optional component registry.

• set-up-document-head

  Use when managing the document head of a frontend SPA — per-route title, meta, Open Graph, and canonical tags via TanStack Router's head option (React) or Unhead (@unhead/vue / @unhead/react), plus a correct html lang. Improves accessibility (titles announced on navigation) and SEO/social previews.

• set-up-error-boundaries

  Use when adding error handling to a frontend app - app-shell, page and component error boundaries with a retryable fallback, a captureError seam, and global handlers, ready for an error tracker.

• set-up-feature-flags

  Use when adding feature flags to a frontend — vendor-agnostic OpenFeature SDK (useFlag hook/composable) with fail-closed defaults, user-targeting context, a dev override, and flag-gated routes. Provider swap (PostHog, GrowthBook, LaunchDarkly) is one line.

• set-up-forms

  Use when adding forms and validation to a frontend project — wires React Hook Form (React) or VeeValidate (Vue) with Zod schema-first validation, where one schema is the single source of truth for shape, rules, and inferred types, with accessible fields and submit wired to a TanStack Query mutation.

• set-up-frontend-structure

  Use when laying down the folder layout of a frontend project - atomic-design components, hooks or composables, libs, utils, stores, styles, and feature modules from the second domain on, with one example component.

• set-up-i18n

  Use when internationalizing a frontend project — sets up i18next + react-i18next (React) or vue-i18n (Vue) with typed message keys, lazy-loaded per-locale catalogs, Intl-based number/date/relative formatting, and a persisted locale choice detected from the browser.

• set-up-motion

  Use when adding animation and transitions to a frontend — uses the native View Transitions API for navigation/layout changes and the Motion library (motion/react, formerly Framer Motion) for component enter/exit/gesture animation, with every animation gated behind prefers-reduced-motion.

• set-up-pwa

  Use when making a frontend installable and offline-capable — sets up vite-plugin-pwa (service worker precaching the app shell, web manifest, auto-update), and optionally persists the TanStack Query cache so cached data is available offline, keeping the SW out of API caching.

• set-up-realtime

  Use when adding live server-push updates to a frontend SPA - a WebSocket seam with backoff reconnect that writes validated pushed data into the TanStack Query cache and exposes connection status as UI state.

• set-up-routing

  Use when adding client-side routing to a frontend SPA — wires TanStack Router (typed, file-based, React) or Vue Router (Vue) with lazy route code-splitting, loaders that prefetch into the TanStack Query cache, per-route error + pending UI, and a guard pattern for protected routes.

• set-up-security-headers

  Use when hardening a frontend SPA's delivery — one strict Content-Security-Policy and header set (no unsafe-inline scripts), rendered for the project's host (Netlify, Vercel, or self-hosted Caddy/nginx/Traefik), plus Dependabot hygiene and an honest note on the XSS surface.

• set-up-state-management

  Use when adding state management to a frontend project - TanStack Query for server state, Zustand (React) or Pinia (Vue) for UI state, a hard boundary between them, a typed query-key factory and a fetch seam.

• validate-env

  Use when a frontend project reads import.meta.env - parses it once with a Zod schema at startup, fails fast with a readable message, and exports a typed env object the seams import instead of raw reads.

• manage-bevy-assets

  Use when a Bevy game loads sprites, audio, levels or custom data — a loading state with handles, one asset-table seam, custom loaders, hot reload in dev, optional processing, and embedded assets so web and single-file builds work.

• optimize-bevy-game

  Use when a Bevy game stutters, drops frames, loads slowly, or ships a large binary or wasm — measure with diagnostics and Tracy/Chrome traces first, then fix change detection, system parallelism, batching, per-frame allocations, and release/wasm profiles.

• scaffold-bevy-game

  Use when starting a new Bevy game or fixing a Bevy project that compiles slowly, has fat code in main.rs, or has no CI — creates a lib+bin cargo project with fast dev builds, plugin composition, assets folder, clippy/rustfmt/nextest CI and optional wasm.

• structure-bevy-app

  Use when adding a feature to a Bevy game or untangling one — plugin-per-feature layout, app states with DespawnOnExit, system sets and ordering, messages vs observers, change detection, query conflicts, and where game logic ends and rendering begins.

• test-bevy-systems

  Use when writing or fixing tests for a Bevy game — headless App/World tests with MinimalPlugins, running one system, asserting messages, observers and state transitions, deterministic time, ordering checks, property tests, and cargo-nextest in CI.

• containerize-service

  Use when a backend or full-stack service needs a production container image — multi-stage Dockerfile per track (Node/pnpm, Go, Python/uv), non-root, digest-pinned bases, health check, SBOM and provenance, image scan.

• deploy-otel-collector

  Use when services emit OpenTelemetry and need one collector per environment — a Compose gateway with memory_limiter, batching, redaction and resource attributes, exporting to SigNoz, grafana/otel-lgtm or any OTLP backend, with a health check and a test span.

• deploy-to-hetzner

  Use when putting a containerized service on Hetzner Cloud with OpenTofu — a VM per environment, Cloud Firewall, private network, hardened cloud-init, Docker Compose with Traefik TLS, deploys by image digest, backups with a restore drill.

• deploy-to-ionos

  Use when hosting on IONOS Cloud with the ionoscloud OpenTofu provider (datacenter, LAN, server or Managed Kubernetes, DBaaS Postgres, Object Storage), or picking IONOS Deploy Now for a static site.

• deploy-to-vercel

  Use when deploying to Vercel - linking the project, env vars per environment and vercel env pull, protected previews, the config file, regions near the database, crons, rolling releases and rollback, the firewall, and gated prebuilt deploys from CI.

• manage-secrets

  Use when a repo handles credentials — decide where each secret lives per host (Vercel, GitHub environments, sops+age for Compose, Supabase), keep them out of images, logs, bundles and state, set up .env.example, rotation and secret scanning.

• set-up-delivery-pipeline

  Use when a repo needs CI/CD on GitHub Actions — PR checks into develop, one image build per commit pushed by digest, dev on develop, stg and prd by approved fast-forward promotion of stage and main — with least-privilege permissions, SHA-pinned actions, OIDC and provenance.

• set-up-opentofu

  Use when infrastructure should be code — OpenTofu repo layout (modules, one root per environment), encrypted remote state with locking, pinned providers, fmt/validate/plan in CI, plan-as-artifact applied after approval, drift detection.

• set-up-terraform-pipeline

  Use when a repo runs Terraform or OpenTofu on GitHub Actions — Azure or AWS via OIDC, develop/stage/main promotion with approval gates, saved plans applied after approval, and weekly drift detection.

• audit-content-quality

  Use when reviewing an article, guide, or landing page before publishing or republishing — scores it against a rubric (project's .claude/rubrics/content-quality.md, else the bundled default) with quoted evidence and fixes only the failed criteria.

• audit-copy-compliance

  Use when page copy must pass brand, legal, or regulatory rules before going live — checks visible copy against a rules file (project's .claude/rubrics/copy-compliance.md, else the bundled template) and reports each violation with quote, rule, and compliant rewrite.

• build-landing-page

  Use when building or restructuring a landing, marketing, or other public conversion page — checks it against a section grammar, one conversion goal, a semantic skeleton, and a hero LCP/CLS budget. Works on built HTML from any stack.

• set-up-lead-capture

  Use when adding or hardening a signup, contact, newsletter, or waitlist form on a public page — form contract, one destination seam, layered spam defense (honeypot, time-trap, Turnstile), GDPR/TDDDG-aware consent, double opt-in.

• set-up-seo

  Use when a public page must be found by search and answer engines — crawlability gate (content in served HTML), metadata, JSON-LD by page type, sitemap, robots.txt incl. AI crawlers, answer-ready structure. Audits built HTML from any stack.

• operate-airtable

  Use when scripting Airtable — scoped PATs, rate limits and 10-record batches, field IDs over names, expiring webhooks, per-environment bases with a schema diff, and tables owned by other systems left read-only.

• operate-aws

  Use when operating AWS by CLI — SSO profiles, explicit profile and region, caller-identity check first, least privilege without long-lived keys, CloudTrail and cost lookups, Session Manager over SSH, CDK version skew avoided.

• operate-azure

  Use when operating Azure by CLI — confirm sign-in, tenant and active subscription first, choose read-only or IaC or a recorded one-off, query with az graph, avoid az rest and RBAC traps, and keep Terraform-managed resources hands-off.

• operate-entra-and-exchange

  Use when managing Entra ID or Exchange Online — app registrations vs enterprise apps, least-privilege Graph permissions, workload identity federation, credential expiry sweeps, shared mailboxes and scoped app access with az ad, az rest and the Graph and Exchange modules.

• set-up-alerting

  Use when a service needs alerts — symptoms users feel, severity P0–P3 with routing, per-environment thresholds, a runbook and owner on every rule; SigNoz is the default (Grafana, Azure Monitor, CloudWatch tracks) and alerts are code promoted with the service through the same pipeline.

Similar plugins