发现 Skills

Anticipate, acknowledge, and respond to reader objections and alternative views in a research argument. Use this skill when the user has a draft argument or storyboard and needs to identify which objections readers will predictably raise, wants to decide which objections to acknowledge and which to set aside, needs vocabulary and sentence templates for introducing and responding to counterarguments without weakening their position, has discovered a flaw in their argument and does not know how to handle it honestly, is building a cause-and-effect argument and needs to address competing causes, has made claims with counterexamples that readers will invoke, uses key terms that readers may define differently and needs to address definitional scope, or wants to avoid either ignoring objections (seeming ignorant) or acknowledging too many (losing argumentative focus). This skill is the detailed companion to research-argument-builder — use it after assembling the argument's core structure (claim, reasons, evidence) and before drafting, to map every acknowledgment slot with a calibrated response strategy.

Apply the complete game-theoretic bargaining framework to any negotiation. Use this skill when a user needs to structure a negotiation, determine who has leverage, calculate the fair split, or decide whether to make a concession or walk away. Triggers include: user is preparing for a salary negotiation, contract renegotiation, partnership deal, M&A term sheet, or labor negotiation and wants to know what number to open with and why; user wants to determine the 'pie' — the true surplus that is actually at stake between the two parties, not the headline dollar figures; user needs to identify and quantify their Best Alternative to a Negotiated Agreement (BATNA) or the other side's BATNA before entering talks; user wants to know how to improve their bargaining position before the negotiation starts (raise your BATNA, lower theirs); user must decide whether to bundle multiple issues together or separate them; user is weighing whether to actually strike, walk out, or threaten to do so, and wants to understand the cost-benefit calculation; user wants to propose a virtual-strike or escrowed-revenue arrangement to eliminate collateral damage while preserving negotiating pressure; user is in an alternating-offer negotiation and wants to calculate the equilibrium split given relative patience levels; user suspects they are negotiating over the wrong number (confusing total value with incremental value above no-deal); user faces brinkmanship — escalating risk of breakdown — and wants to calibrate how far to push. This skill does NOT cover simultaneous-move games (use nash-equilibrium-analyzer), one-shot ultimatum games without iteration, or multi-party coalition bargaining beyond two principal parties.

Test web application back-end components for non-SQL server-side injection vulnerabilities. Use this skill when: testing for OS command injection via shell metacharacters (pipe, ampersand, semicolon, backtick) or dynamic execution functions (eval/exec/Execute); detecting blind command injection using time-delay technique (ping -i 30 loopback) when output is not reflected; probing for path traversal vulnerabilities including filter bypass via URL encoding, double encoding, 16-bit Unicode, overlong UTF-8, null byte injection, or non-recursive strip bypass; testing for Local File Inclusion or Remote File Inclusion; identifying XML External Entity (XXE) injection for local file read or Server-Side Request Forgery (SSRF); detecting SOAP injection via XML metacharacter probing; testing for HTTP Parameter Injection (HPI) and HTTP Parameter Pollution (HPP) in back-end HTTP requests; identifying SMTP injection through email header manipulation or SMTP command injection in mail submission forms. Covers detection procedures, filter bypass techniques, exploitation impact, and prevention countermeasures. Maps to CWE-78 (OS Command Injection), CWE-22 (Path Traversal), CWE-98 (File Inclusion), CWE-611 (XXE), CWE-91 (XML Injection), CWE-88 (Argument Injection), CWE-93 (SMTP Injection). For authorized security testing, security code review, and defensive hardening contexts.

Answer Engine Optimization (AEO): optimize content to be cited by LLMs (ChatGPT, Claude, Perplexity, Gemini, Copilot) when they answer user questions. Use when designing content strategy for the LLM-citation era, auditing existing content for LLM citability, structuring Q&A schema for AI surfaces, tracking which content gets cited and where, or measuring AEO performance against competitors. Distinct from our `ai-seo` skill — that one is about ranking in AI search results (Perplexity, Google AI Overviews); this one is about being the source LLMs cite in their answers.

Epic decomposition into trackable issues. Triggers: "create a plan", "plan implementation", "break down into tasks", "decompose into features", "create beads issues from research", "what issues should we create", "plan out the work".

name: gaap-standards description: Key US GAAP standards — ASC 606, 842, 350, 805

Design and document visual identity systems including color, typography, imagery, and layout principles.

Syncs active Claude Code plugins (skills and agents) to Copilot CLI. Use this skill at session start, when the user mentions Claude plugins, when skills or agents seem missing, when the user asks about available capabilities, or when troubleshooting why a skill isn't working.

IoT network traffic analyzer for detecting IoT protocols and identifying security vulnerabilities in network communications. Use when you need to analyze network traffic, identify IoT protocols, or assess network security of IoT devices.

Fetch Xiaohongshu (RedNote / xhs) note detail and comments by note ID, returning title, description, author info, engagement stats, tags, and a paginated comment list. Use when a user mentions note detail, Xiaohongshu, RedNote post, xhs note content, Xiaohongshu comment scrape, fetch post comments, scrape xhs comments, RedNote note details, Xiaohongshu post engagement, Xiaohongshu note content, RedNote post data, xhs post detail, get Xiaohongshu comments, RedNote comment list, Xiaohongshu note metadata, xhs note scrape, or fetch RedNote note.

Toolkit for styling artifacts with a theme. These artifacts can be slides, docs, reportings, HTML landing pages, etc. There are 10 pre-set themes with colors/fonts that you can apply to any artifact that has been creating, or can generate a new theme on-the-fly.

Audit a multi-tab or multi-page web app for duplicate visuals, charts, or components. Map what appears on each tab, identify duplicates, decide ownership, and remove redundant instances surgically. Typical target: dashboards with tabbed navigation where the same chart or metric card appears on multiple tabs.

用于 FlowGram 物料库组件文档撰写的专用技能，提供组件文档生成、Story 创建、翻译等功能的指导和自动化支持

Read and summarize posts from Telegram channels via MTProto (Pyrogram or Telethon). Fetch recent messages from public or private channels by time window.

Initialize a dagster project using the create-dagster cli. Create a dagster project, uv virtual environment, and everything needed for a user to run dg dev or dg check defs successfully. (project)

Validate the entire Claude Code configuration is working correctly.

Comprehensive guide for BFL FLUX image generation models. Covers prompting, T2I, I2I, structured JSON, hex colors, typography, multi-reference editing, and model-specific best practices for FLUX.2 and FLUX.1 families.

Diagnose and fix bugs through systematic investigation, root cause analysis, and targeted validation. Use when something is broken, errors occur, performance degrades, or unexpected behavior manifests.

Reduces token usage from paid providers by offloading work to local LM Studio models. Use when: (1) Cutting costs—use local models for summarization, extraction, classification, rewriting, first-pass review, and brainstorming when local quality suffices, (2) Avoiding paid API calls for high-volume or repetitive tasks, (3) No extra model configuration—just-in-time model loading and REST API integration work with your existing LM Studio setup, (4) Local-only or privacy-sensitive work. Requires LM Studio 0.4+ with server (default :1234). No CLI required.

Create a purchase claim (PCL) for price or quantity discrepancies with a supplier

Clean all build artifacts and rebuild from scratch. Use when experiencing stale build artifacts, Turbo cache issues, mysterious build failures, or after dependency updates.

Renaming a C file or associated header (.h) file requires additional steps in decompilation projects or the build will fail. Use this skill when you want to rename a C file and/or a header file.

Consult official Claude Code documentation from docs.claude.com using selective fetching. Use this skill when working on Claude Code hooks, skills, subagents, MCP servers, or any Claude Code feature that requires referencing official documentation for accurate implementation. Fetches only the specific documentation needed rather than loading all docs upfront.

Designs and blueprints intentionally designed outdoor space in a live Minecraft Bedrock world — formal gardens, parks, plazas, courtyards, hedge mazes, fountains, parterres, topiary, cloister gardens, estate grounds. Covers French formal, Italian, English landscape, Mughal, Japanese, Chinese, modernist, and other traditions. Geometric and intentional — the deliberate counterpart to the naturalistic terraforming skill. Use when the user wants a garden, park, plaza, courtyard, hedge maze, fountain, or designed grounds. Part of the minecraft-builder workflow.