Explore skills

A test skill for validating the packager

End-to-end Playwright skill pack for planning, authoring, debugging, documenting, and operationalizing test automation.

This skill analyzes your Douban reading data, generates reports, visualizes your reading taste profile, and provides personalized insights. It collects all data from your account (including non-public) to create an HTML report and AI-driven insights, not for general recommendations or crawling others' pages.

Accessibility engineering skill for web products with JAWS and NVDA testing plus Spain/EU compliance guidance. Use when the task involves WCAG 2.2 implementation, ARIA behavior, keyboard/focus issues, screen-reader regressions, accessibility audits, remediation plans, or legal scope checks for Ley 11/2023, RD 1112/2018, EAA, and EN 301 549.

Turn Gmail into structured dashboards — extract and track subscriptions, packages, invoices, job applications, and more.

A Gentle 'Big Sister' Skill featuring an AI persona that is soft on the outside and strong on the inside, providing powerful, inclusive, and healing emotional support.

Use when test files are accumulating across projects, test suites are growing unwieldy, flaky tests appear, test execution is slow, user asks about test cleanup or archival strategy, or when organizing tests after skill-generated workflows like bkit or superpowers TDD cycles.

Configures Claude Code to launch third-party LLM backends (GLM, DeepSeek, Qwen, Kimi, etc.) via their Anthropic-compatible endpoints. It creates per-provider launcher scripts and shell aliases, allowing users to start Claude Code with a specific backend using simple commands like glm or qwen.

Scans a codebase to identify hand-rolled implementations that should be replaced by third-party libraries, missing capabilities, and code organization issues (directory structure, naming, circular dependencies, barrel bloat). Produces structured migration plans with Context7-verified recommendations.

Ice Queen/Dominatrix Skill with an imposing aura and decisive actions. Serves as a strong, clear-headed AI decision mentor.

Universal Java coverage backfill workflow for any Java project. It detects the build tool, adds JaCoCo if missing, analyzes low-coverage classes, adds targeted tests, and iterates with CLASS/METHOD/LINE/BRANCH metrics until the threshold is met or blockers are reported.

A Claude Code AI Persona for rational decision-making, stopping losses, and wealth creation, designed to counter emotional dependency and mental burnout.