FujishigeTemma

Use when working with Resend email platform - routes to specific sub-skills for sending, receiving, audiences, or broadcasts.

Best practices for building applications with React 19's Async Transition model combined with TanStack Router and TanStack Query. Covers the "everything is a Transition" paradigm, Suspense architecture, component design patterns (Boundary vs Leaf components, Action Props, AsyncBoundary), state ownership decisions (server state vs URL state vs ephemeral state), useTransition with async functions, useOptimistic, data-pending visual feedback, and file-based routing directory structure with co-location. Use this skill when designing component architecture, deciding where state should live, structuring a TanStack Router project, building design system components that handle async operations, or writing any code that involves useTransition, Suspense boundaries, useOptimistic, or startTransition. This complements tanstack-query-best-practices (which covers Query-specific APIs like queryOptions, query key factories, mutations, and caching) by focusing on the broader React async architecture that Query operates within

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

Comprehensive best practices for writing TanStack Query (React Query) v5 code, based on all 31 posts from TkDodo's Practical React Query blog series. Covers query key factories with queryOptions, mutations and invalidation strategies, TypeScript type inference, error handling, data transformations with select, status checks, testing, optimistic updates, and architectural patterns. Use this skill whenever writing, reviewing, or refactoring code that uses TanStack Query — including useQuery, useMutation, queryOptions, queryClient, invalidation, prefetching, or any @tanstack/react-query import. Also use when the user asks about React Query patterns, caching strategies, or server state management. This applies even for small changes like adding a single query hook.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Table-driven test pattern for writing clean, maintainable, and comprehensive tests. Separate test data (table entries) from test logic to eliminate copy-paste duplication. Use when: (1) writing Go tests with multiple input/output variations, (2) testing pure functions with various combinations, (3) refactoring duplicated test code, (4) adding edge cases and boundary conditions to existing tests, (5) implementing TDD test scenarios that share the same assertion structure. Covers Go-specific patterns (t.Run subtests, parallel execution, cmp.Diff) and language-agnostic principles applicable to any parameterized testing.

Test-Driven Development workflow based on t_wada and Kent Beck's Canon TDD. Enforce the List-Red-Green-Refactor cycle for AI agent development. Use when: (1) implementing features with TDD, (2) AI agent is delegated to write code with tests, (3) user requests TDD or test-first development, (4) ensuring correctness of AI-generated code through test-driven discipline.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Strict TDD workflow with property-based testing using pytest and hypothesis. Use when writing new tests, updating tests, or when asked to test code. Enforces t_wada-style Red-Green-Refactor discipline: one test at a time, confirm failure before implementation. Covers property identification, hypothesis strategy design, and test design techniques (boundary value analysis, equivalence partitioning).

Add, revise, and review tests for the edmkit repository with repo-specific guidance for pytest, Hypothesis, numerical tolerances, synthetic dynamical systems, slow or gpu markers, and smoke-test constraints. Use when Codex is changing code under src/edmkit/, touching tests/, or needs to decide the right test strategy for embedding, simplex projection, S-Map, CCM, metrics, splits, or generators in this repository.

Design, implement, review, and modernize Python tests with an emphasis on pytest, Hypothesis, and AI-era evaluation workflows. Use when Codex needs to add or fix tests, improve flaky or slow suites, review pytest structure and fixture usage, introduce property-based testing, or separate deterministic code tests from eval-style checks for LLM or agent behavior.

Guide use of the `semantic-scholar` CLI (installed via uvx) for read-only Semantic Scholar research workflows. Use when the agent needs to search papers, inspect paper metadata, follow citations or references, look up authors, or export formatted citations. Prefer this skill when a task involves literature discovery, paper disambiguation, author disambiguation, or citation lookup against Semantic Scholar. Do not use it for general web search, non-Semantic-Scholar sources, or repository code tasks unrelated to academic literature.

Opinionated guidance for building and refactoring Vue 3.5 TypeScript SPAs that use Vue Router v5 built-in file-based routing, Router Data Loaders, Vue Suspense, and TanStack Query Vue v5. Use when Codex needs to implement or review data fetching, mutations, navigation, URL query params, route data loaders, Suspense, async setup, optimistic UI, error handling, or testing in a non-SSR CSR app and must enforce consistent async UI contracts instead of offering multiple styles.