发现 Skills

Audit inline complexity and recommend variable extraction; produces a report with flattening suggestions for nested expressions.

Convert HuggingFace transformer models to ONNX format for browser inference with Transformers.js and WebGPU. Use when given a HuggingFace model link to convert to ONNX, when setting up optimum-cli for ONNX export, when quantizing models (fp16, q8, q4) for web deployment, when configuring Transformers.js with WebGPU acceleration, or when troubleshooting ONNX conversion errors. Triggers on mentions of ONNX conversion, Transformers.js, WebGPU inference, optimum export, model quantization for browser, or running ML models in the browser.

Web page archiving and retrieval from cached/deleted sources. Use when accessing unavailable pages, preserving web content, creating legal evidence archives, or building redundant archival workflows. Covers Wayback Machine, Archive.today, ArchiveBox, and evidence preservation tools.

Open Source Intelligence gathering and attack surface management for external reconnaissance.

Use when building MCP servers or clients that connect AI systems with external tools and data sources. Invoke for MCP protocol compliance, TypeScript/Python SDKs, resource providers, tool functions.

Calculate cryptocurrency tax obligations with cost basis tracking, capital gains computation, and Form 8949 generation. Use when calculating crypto taxes, generating tax reports, comparing cost basis methods, or identifying taxable events. Trigger with phrases like "calculate crypto taxes", "generate tax report", "cost basis FIFO", "capital gains", "Form 8949", or "crypto taxable events".

Build discounted cash flow (DCF) valuation models in Excel with free cash flow projections, WACC calculations, and sensitivity analysis for investment banking and corporate finance teams

Install and configure TwinMind Chrome extension, mobile app, and API access. Use when setting up TwinMind for meeting transcription, configuring calendar integration, or initializing TwinMind in your workflow. Trigger with phrases like "install twinmind", "setup twinmind", "twinmind auth", "configure twinmind", "twinmind chrome extension".

Build browser automation and web scraping with Playwright on your local machine. Prevents 10 documented errors including CI timeout hangs, extension testing failures, and Ubuntu compatibility issues. Includes stealth mode for anti-bot bypass, authenticated sessions, infinite scroll handling, screenshot/PDF generation, and v1.57 Speedboard performance analysis. Use when: automating browsers, scraping protected sites, testing with real IPs, bypassing bot detection, generating screenshots/PDFs, or troubleshooting "target closed", "page.pause() hangs CI", "permission prompts block tests", or "Ubuntu 25.10 installation" errors.

Write comprehensive clinical reports including case reports (CARE guidelines), diagnostic reports (radiology/pathology/lab), clinical trial reports (ICH-E3, SAE, CSR), and patient documentation (SOAP, H&P, discharge summaries). Full support with templates, regulatory compliance (HIPAA, FDA, ICH-GCP), and validation tools.

Create research posters using HTML/CSS that can be exported to PDF or PPTX. Use this skill ONLY when the user explicitly requests PowerPoint/PPTX poster format. For standard research posters, use latex-posters instead. This skill provides modern web-based poster design with responsive layouts and easy visual integration.

Create publication-quality scientific diagrams using Nano Banana 2 AI with smart iterative refinement. Uses Gemini 3.1 Pro Preview for quality review. Only regenerates if quality is below threshold for your document type. Specialized in neural network architectures, system diagrams, flowcharts, biological pathways, and complex scientific visualizations.

IRAC legal-research skill for Taiwan in-house legal teams. Accepts a legal question / statute citation / case query and runs a plan-first semi-interactive Agent loop (LLM drafts a search plan → user Y/n confirms → autonomous WebFetch + triangulation + Harvey document-level citation), producing a 4-file output (plan.md + state.json + research-memo.md + executive-summary.md). Loop cap: ≤ 5 rounds OR ≤ 30 fetches; early stop if ≥ 8 sources AND ≥ 2 source types; forced_stop emits a ⚠️ marker. No profile.yml dependency. §6.3 disclaimer footer + §6.4 escalation override inherited from v0.4.x. TRIGGER: - "查 §" / "§227 是" / "判例" / "法條" / "找判決" / "research" / "lookup" / "條文" / "函釋" - "law-text lookup" / "find precedent" / "research the case" / "what does §X say" - statute number + code name patterns (e.g. "Civil Code §184" / "Personal Data Protection Act §27") USE WHEN: the user wants literal law text / precedent / official interpretation lookup with structured iterative search + triangulation + citation. NOT for fact-pattern legal analysis (use legal-issue-spot).

Styles applications using Tailwind CSS v4 utilities. Activates when adding styles, restyling components, working with gradients, spacing, layout, flex, grid, responsive design, dark mode, colors, typography, or borders; or when the user mentions CSS, styling, classes, Tailwind, restyle, hero section, cards, buttons, or any visual/UI changes.

Create and manage scheduled tasks (cron jobs) that send you messages at specified times.

Send and read iMessages/SMS from macOS. Use for texting contacts, scheduling services, or automating message-based workflows. Triggers on queries about texting, messaging, SMS, iMessage, or contacting someone via text.

Guides the creation of self-signed SSL/TLS certificates using OpenSSL, including key generation, certificate creation, combined PEM files, and verification scripts. This skill should be used when tasks involve generating self-signed certificates, creating SSL certificate infrastructure, or writing certificate verification scripts.

Digital forensic skill for recovering passwords and sensitive data from disk images, deleted files, and binary data. This skill should be used when tasks involve extracting passwords from disk images, recovering deleted file contents, analyzing binary files for fragments, or forensic data recovery scenarios. Applies to tasks mentioning disk images, deleted files, password fragments, or data recovery.

Search Yelp for local businesses, get contact info, ratings, and hours. Use when finding services (cleaners, groomers, restaurants, etc.), looking up business phone numbers to text, or checking ratings before booking. Triggers on queries about finding businesses, restaurants, services, or 'look up on Yelp'.

Internal skill - Research subagent that executes focused research tasks using web tools. Called automatically by the deep-research lead agent.

Guide for using TB2J command-line tools to calculate magnetic interaction parameters (J, DMI, etc.) from DFT outputs (Wannier90, Siesta, Abacus). Use this skill when the user asks about running TB2J commands, their parameters, inputs, or outputs.

Use this skill when a user invokes /novel-core or requests writing long-form web fiction, starting a book, planning plot, writing chapters, proofreading, removing an "AI" tone, humanizing/polishing prose, or analyzing genre references. This skill is the entry point for the novel-harness project: it will load .harness/agents/总编Agent.md and, according to the routing under .harness/agents/, dispatch tasks to planning, writing, proofreading, and context Agents.

React 18+ patterns with the Vite bundler, TypeScript, hooks, component design, and state management. Activate this Skill for React apps, Vite configuration, TypeScript typings, frontend development, and modern React patterns.

Guide for creating evolving skills — detailed workflow plans that capture what you'll do, what tools you'll create, and lessons learned from execution. Use this when starting a new task that could benefit from a reusable workflow.