发现 Skills

Comprehensive testing patterns and anti-patterns for writing and reviewing tests

Karpathy's LLM Wiki — build and maintain a persistent, interlinked markdown knowledge base with human feedback loop. Ingest sources, query compiled knowledge, lint for consistency, and process audit feedback.

Full patent drafting pipeline from an invention description to jurisdiction-formatted filing documents. Supports CN (CNIPA), US (USPTO), and EP (EPO). Supports invention patents and utility models. Use this when a user says "write a patent", "patent pipeline", "patent application", "draft patent", "write claims", or wants to draft a complete patent application.

Full research pipeline: Workflow 1 (idea discovery) → implementation → Workflow 2 (auto review loop). Moves from a broad research direction all the way to a submission-ready paper. Use when the user says "full pipeline", "from idea discovery to submission", "end-to-end research", or wants the complete autonomous research lifecycle.

Apply Benjamin Graham's value investing framework to evaluate stocks, portfolio allocation, and investment vs. speculation decisions. Trigger on: 'Is this stock worth buying?', 'Is this investment or speculation?', 'How should I allocate my portfolio?', 'Is this company a good value?', 'should I sell in a downturn?', 'evaluate this stock for a defensive investor'.

Expert photography prompt engineer specializing in crafting detailed, evocative prompts for AI image generation. Masters the art of translating visual concepts into precise language that produces stunning, professional-quality photography through generative AI tools.

AI filter remover — remove photo filters and restore natural image colors

Generate and maintain AI-friendly project architecture documentation, complete with Mermaid diagrams for architecture, data flows, sequences, and state. Use this skill whenever the user asks to scaffold project docs, create an ARCHITECTURE.md, AGENTS.md, design docs, execution plans, product specs, or any kind of structured project documentation meant to give an AI (or human) full context on a codebase. The skill scaffolds docs that don't exist yet and updates docs that already do. Also trigger when the user asks to audit, update, regenerate, or add diagrams to existing project documentation, or when they mention "project docs", "architecture docs", "documentation structure", "document my project", "codebase docs", "AI context docs", "architecture diagram", "mermaid diagram", "flow diagram", or similar. Even if they just say "set up docs" or "I need docs for my project" — use this skill.

Analyze disk space and find cleanable files. Use when the user wants to clean a disk, free up space, find large files, analyze disk usage, or asks what’s taking up space. Triggers on requests like "clean my disk", "free up space on C:\\", "find large files", "what's using my disk space", "disk cleanup".

This skill should be used when the user asks about "prompt_toolkit Buffer", "Document", "cursor position", "text manipulation", "BufferControl", "text editing", "undo/redo", "selection", "copy paste", or needs to handle text input and manipulation in prompt_toolkit applications.

Develop, optimize, and troubleshoot CrowdStrike LogScale (Humio) security detection queries using CQL syntax. Use this when writing LogScale queries, building security detections, creating threat-hunting rules, fixing CQL syntax errors, working with CrowdStrike EDR/Falcon security monitoring, or building behavioral rules with the correlate() function. Handles case statements, risk categorization, multi-event correlation, investigation playbooks, and actionable security outputs.

아이디어 기획서 또는 서비스 설명을 기반으로 가상 페르소나를 생성하고, AI 인터뷰를 시뮬레이션하여 사용자 인사이트를 도출합니다.

코딩, 리뷰, 리팩터링, 디버깅, 구현 계획에서 Karpathy식 coding agent discipline을 적용해 추측, 과설계, 주변 변경, 약한 검증을 줄일 때 사용한다.

Use to audit test quality with Google Fellow SRE-level scrutiny — identifies tautological tests, coverage-gaming patterns, weak or brittle assertions, and missing corner cases. Creates a backlog epic with prioritized tasks for improvements, then runs SRE-style task refinement on each item.

Use when errors occur deep in execution — traces bugs backward through the call stack to find the original trigger, not just the symptom

Run tests, linting, and quality checks for WooCommerce development. Use when running tests, fixing code style, or following the development workflow.

Managing Linear issues, projects, and teams. Use when working with Linear tasks, creating issues, updating status, querying projects, or managing team workflows.

Generate forms with Form.* components and validation. Use when creating input forms, multi-step wizards, or form validation.

Create distinctive, production-grade frontend interfaces that avoid generic AI aesthetics. Use when building web components, pages, dashboards, or styling web UI.

专为Dify工作流案例分享设计的公众号文章生成器,遵循小灰灰公众号写作规范,自动生成包含前言、工作流制作、总结的完整Dify案例文章,配有详细的节点配置、插件安装步骤、代码示例,并支持自动生成图片上传到腾讯云COS图床

TypeScript best practices and patterns for writing clean, type-safe, and maintainable code. Use when implementing TypeScript code, refactoring for better type safety, or teaching TypeScript patterns.

Orchestrate parallel scientist agents for comprehensive research with AUTO mode

Refactor high-complexity React components in Dify frontend. Use when `pnpm analyze-component --json` shows complexity > 50 or lineCount > 300, when the user asks for code splitting, hook extraction, or complexity reduction, or when `pnpm analyze-component` warns to refactor before testing; avoid for simple/well-structured components, third-party wrappers, or when the user explicitly wants testing without refactoring.

Replaces ALL built-in search tools. You MUST invoke this skill BEFORE using WebSearch, Grep, or Glob. NEVER use the built-in Grep tool - use `grepai` instead.