Plugin

Use this skill when migrating between technologies (REST → gRPC, monolith → microservices, on-prem → cloud, framework upgrade), executing a major redesign driven by tech debt or scale, or a strategic platform shift (database engine, message bus, identity provider, observability vendor) — including when the user describes a "from X to Y" move without saying "evolve" or "migration" — to plan and document architecture migration and evolution — an ADR for the migration decision, target-state ARCHITECTURE.md, a phased migration plan (strangler fig, expand-contract, branch by abstraction), fitness functions, and a rollback strategy. Not for new feature design — use `/architecture-design` instead.

Use this skill when designing AI agent systems or RAG pipelines that need a principled approach to selecting, structuring, and delivering context to LLMs — the context engineering knowledge base covering the context stack model, RAG pipeline design, memory engineering, and multi-agent context coordination.

Use this skill when bootstrapping scheduled knowledge-base sync for a repo that has no knowledge/.knowledge-sync.yml yet — to run one-time setup that detects the knowledge_root from CLAUDE.md/AGENTS.md, maps doc areas to source globs, records opt-in external sources (Linear/Notion/WebFetch, all disabled by default), captures a baseline last_scanned_sha, sets the per-area update policy, generates or seeds knowledge/CONVENTIONS.md, provisions the L4 memory dir, and offers to register the daily routine. Routes ongoing recurring sync operations to /knowledge-sync.

Use this skill when scaffolding a new skill (workflow, knowledge, or companion utility) under `plugin/skills/` and you need spec-conformant frontmatter, an eval test-case stub, and memory-write points pre-wired — the internal procedure for `/plugin-author create`. Generates a new skill INSIDE the ai-skills plugin conforming to the agentskills.io specification (https://agentskills.io/specification) and ai-skills plugin conventions. No longer slash-invocable — call `/plugin-author create <name>` instead. Read by the `prompt-engineer` agent at task start when DEV-ing or reviewing a plugin skill.

Use this skill when the agent is implementing Java backend services in Spring Boot projects, reviewing JPA entity models, configuring connection pools or caches, designing REST endpoints, writing repository/service/controller layers, or onboarding to Spring conventions — for Spring Boot 3 + JPA backend patterns covering Java 21+ features, layered architecture, Spring Data JPA, Hibernate fetch strategies, Flyway migrations, HikariCP pool tuning, Redis caching with Spring Cache, Spring Security 6 filter chains, REST API design (RFC 7807, pagination, versioning), Testcontainers + JUnit 5 + Mockito testing, Micrometer + Actuator observability, and virtual threads and performance tuning.

Use this skill when a service, module, or system needs documentation, audit, or tech-debt assessment — onboarding a new team to an undocumented service, establishing a pre-redesign "what do we actually have?" baseline, or a tech-debt audit before roadmap planning, including when the user does not say "architecture" — to analyze the existing architecture and produce ARCHITECTURE.md, current-state C4 diagrams, a component inventory, gap analysis (current vs target), and a technical-debt register. Routes to system-architect (default — system/module/service analysis), cloud-architect (`--cloud` for cloud-resource inventory + cost + compliance review), or devops-architect (`--cicd` for CI/CD platform audit + DORA baseline + governance review).

Use this skill when the user asks to draft, write, or improve tweets, threads, LinkedIn posts, LinkedIn newsletter issues, Facebook posts, Threads posts, Bluesky posts, promote a blog post, share on social, or repurpose content for social channels — to create social media posts across X/Twitter, LinkedIn (posts + newsletters), Facebook, Threads, and Bluesky, reading brand context from marketing/MARKETING.md at runtime.

Use this skill when authoring or running a Terraform plan/apply, refactoring modules, performing state surgery, importing existing cloud resources, or working in repos that use OpenTofu instead of HashiCorp Terraform — Terraform and OpenTofu procedures covering state operations (state list/show/mv/rm/pull/push), plan/apply lifecycle, OpenTofu fork compatibility, state-surgery safety warnings, and workspace handling. Loaded by `/infra-change` and other infrastructure workflows; not a workflow itself.

Use this skill when authoring or reviewing a deploy workflow that needs the strategy and rollback knowledge — the deployment strategy reference (rolling, blue-green, canary, recreate) plus rollback decision criteria, pre-deploy checklist, and post-deploy verification, loaded by the `/deploy-production` and `/deploy-staging` workflows; not a workflow itself.

Use this skill when picking a diagnostic vocabulary for a latency, error-rate, saturation, or crashloop investigation, when authoring or reviewing an analysis workflow that needs the method reference, or when correlating SLI metrics and active alerts against a method's signal set — a knowledge skill of industry-canonical observability methodologies (Four Golden Signals, RED, USE, Distributed Tracing) with method-to-problem mapping that explains which signals each method surfaces and when each method applies. Loaded by `/analyze-prod`, `/analyze-local`, `/env-analyze`, `/infra-change`, and `/bugfix` workflows when production-context diagnosis needs a named methodology.

Use this skill when the agent is implementing frontend code in Next.js projects, reviewing React or Next.js pull requests, onboarding to App Router conventions, or designing a frontend architecture that needs canonical React/Next.js patterns — for canonical React and Next.js conventions covering Next.js App Router routing and rendering, React Server Components vs Client Components boundary, TypeScript strict-mode discipline, Tailwind CSS and shadcn/ui styling, React Query and Server Actions data fetching, Core Web Vitals performance optimization, WCAG 2.2 accessibility, Next.js metadata API and structured data for SEO, unit testing with Vitest and React Testing Library, and end-to-end testing with Playwright.

Use this skill when the agent is designing schemas, reviewing migrations, tuning queries, modeling NoSQL access patterns, configuring replication, planning partitioning, auditing indexes, or onboarding to database conventions — the database patterns knowledge base for PostgreSQL, MySQL, SQL Server, MongoDB, Redis, Cassandra, DynamoDB engines, schema design, SQL optimization, indexing strategies (B-tree, hash, GIN, GiST, BRIN, partial, covering), partitioning (range, list, hash), replication topologies, NoSQL data modeling, time-series, graph, monitoring, capacity planning, backup and recovery.

Use this skill when the agent is running a security audit, scanning dependencies, prioritizing CVEs by exploit likelihood, signing release artifacts, generating SBOMs for compliance, or reviewing CI/CD artifact integrity — software supply chain security knowledge covering SBOM generation (Syft, CycloneDX, SPDX), SLSA provenance levels 1 through 4, Sigstore Cosign artifact signing, EPSS exploit-prediction scores, CISA KEV known-exploited catalog, dependency-confusion / typosquatting / lockfile-validation patterns, and EO 14028 and EU Cyber Resilience Act SBOM mandates.

Use this skill when running tests in a specific stack, wiring CI test steps, or designing test infrastructure — per-stack test runner commands and conventions for TypeScript/JavaScript, Java, Python, Go, and .NET covering unit, integration, E2E, and coverage commands plus Testcontainers conventions and quality-gate signals, a reference table for picking the right command after stack detection. Knowledge skill — loaded as context by `/test-local` and `/test-strategy`, never invoked directly.

Use this skill when designing UI, defining or auditing a design system, reviewing UX deliverables, specifying components, evaluating accessibility, or producing redlines and tokens for engineering handoff — the UI/UX knowledge base covering user research, personas, jobs-to-be-done, user journeys, usability testing, information architecture, wireframing, prototyping, visual design, typography, color theory, grid systems, visual hierarchy, micro-interactions, animation, design systems, atomic design, design tokens, Figma workflows, WCAG 2.2 accessibility, responsive and mobile-first design, Material Design 3, Apple HIG, Nielsen heuristics, Gestalt principles, conversion-centered design, and design-to-code handoff.

Use this skill when the agent is picking a release tool for a stack or running version bump + changelog generation inside the `/release` workflow — for the release tooling reference covering release-please, semantic-release, Changesets, GoReleaser, cargo-release, JReleaser, plus monorepo per-package versioning via Changesets, Lerna, and Nx, including detection markers, per-tool version-bump + changelog commands, and per-stack release conventions for Node, Go, Rust, and Java.

Use this skill when designing or implementing Python backend code in FastAPI projects, reviewing Python pull requests, or onboarding new contributors to FastAPI conventions — a Python backend patterns knowledge base covering Python 3.12+, FastAPI layered architecture, Pydantic v2 schemas, SQLAlchemy 2.0 async ORM, Alembic migrations, async/await, dependency injection, pytest with httpx.AsyncClient, OAuth2/JWT security, ruff and mypy strict, structured logging and Prometheus metrics.