appkit

MUST invoke when generating mn.* based UIs or the user asks how to use, configure, or debug any mn.* or appkit_mantine component. appkit_mantine is NOT in Claude's training data — always use this skill rather than guessing at APIs. Covers building forms, tables, drawers, modals, tooltips, color pickers, date/time pickers, charts, and navigation; looking up correct props; debugging on_change handler types (NumberInput sends float|str, DateInput sends str|""); troubleshooting component quirks like tooltips on disabled elements, overlays not closing, or props having no effect.

Expert guide for runic.ogm — a SQLModel-style, graph-native Python OGM for Cypher databases (FalkorDB, Neo4j, Memgraph, ArcadeDB, Apache AGE). Use whenever the user defines graph models (Node/Edge), maps fields, declares relationships, writes graph queries/traversals, or does session/repository CRUD with runic. Invoke for any code that imports from `runic.ogm`, uses `Node`, `Edge`, `Field`, `Relation`, `Session`, `Repository`, `select()`, or the query builder, and for any "how do I model/query this graph in runic" task. This is the OGM skill; for schema migrations use the `runic-migrate` skill instead.

appkit-commons usage patterns: configuration (YAML profiles, env overrides, secrets), service registry, repository pattern, database entities, custom column types (EncryptedString, ArrayType), and scheduler (APScheduler, PGQueuer). Apply automatically when adding new features, services, repositories, scheduled tasks, or configuring the application stack.

Use when the user wants to refine, sharpen, or expand a rough idea into a detailed implementation prompt — before any code gets written. Trigger on /boost, "refine this prompt", "turn this into a prompt", "help me specify", "what do I need to think through before implementing", "can you make this more detailed", rough feature ideas the user explicitly wants help scoping (not implementing). Also trigger when user says "boost this" or hands you a vague feature description and asks for a proper prompt or spec. Do NOT trigger for direct implementation requests, debugging questions, code reviews, or architectural questions where the user is not asking for a refined prompt.

Refactor and simplify Python files modified in the current session — minimal diff, clean code principles, then lint/format/typecheck. Use when the user asks to clean up, simplify, or refactor code.

Generate a conventional commit message from staged (or unstaged) changes. Use when the user asks for a commit message, wants to commit changes, or says "what should my commit message be".

Create a comprehensive, well-structured README.md for the project. Use when asked to write, generate, or update a project README. Trigger on "create readme", "write readme", "generate readme", or /readme.

Creates optimized multi-stage Dockerfiles for any language or framework. Use when building, reviewing, or refactoring Dockerfiles, container images, or Docker Compose configurations. Covers stage structure, base image selection, layer caching, security hardening, and healthchecks.

Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, artifacts, posters, or applications (examples include websites, landing pages, dashboards, reflex.dev components, HTML/CSS layouts, or when styling/beautifying any web UI). Generates creative, polished code and UI design that avoids generic AI aesthetics.

Create, audit, or troubleshoot GitHub Actions CI/CD pipelines. Use when: building a new workflow, reviewing an existing pipeline for security/performance, debugging a failing workflow, designing deployment strategies, or setting up secrets/OIDC authentication. Triggers: "create workflow", "CI/CD pipeline", "GitHub Actions", "review pipeline", "audit workflow", "fix workflow", "deployment strategy", "OIDC", "caching", "matrix strategy".

Enforce Clean Code Developer (CCD) architecture and software quality principles when writing or refactoring Python code. Apply automatically whenever you generate new Python code, refactor existing Python, design modules or class hierarchies, or review Python architecture — even when the user doesn't mention "clean code" or "architecture" explicitly. Covers all four CCD grades: Orange (SRP, SoC, SLA, conventions), Yellow (ISP, DIP, LSP, information hiding, least astonishment), Green (OCP, Tell Don't Ask, Law of Demeter), Blue (YAGNI, design-first, component orientation). Triggers on: "write a class", "add a service", "refactor this", "how should I structure", "implement feature X", and any Python file editing task.

Enforces Python 3.14 code style, logging rules, type annotations, design patterns, testing strategy, and clean code principles for the appkit project. Apply automatically whenever writing, reviewing, or refactoring Python code, creating new modules, or adding tests.

Reflex state design, event handlers, background tasks, form validation, page factory, service registry, repository pattern, database models, project architecture.

Bump version, generate an AI-written changelog, and publish a GitHub release. Use when the user says "release", "publish", or "bump version".

Expert guide for runic — an Alembic-style graph schema migration tool that supports FalkorDB, Memgraph, Neo4j, ArcadeDB, and Apache AGE. Use whenever the user works with runic: CLI commands, migration file anatomy, op.* API, env.py config, autogenerate, branching/merge, or testing. Invoke for any versions/*.py file, `runic` CLI question, or graph schema change task.

Azure Verified Modules (AVM) requirements and best practices for developing certified Azure Terraform modules. Use when creating or reviewing Azure modules that need AVM certification.

Transform monolithic Terraform configurations into reusable, maintainable modules following HashiCorp's module design principles and community best practices.

Guide for running acceptance tests for a Terraform provider. Use this when asked to run an acceptance test or to run a test with the prefix `TestAcc`.