Top Agent Skills

AI 论文速读 Skill：三档深度（裸读 / 引导 / 精读）+ 页码级 Provenance 防幻觉 + 多篇对比。 触发词：论文速读、读这篇论文、抓核心观点、论文对比、多篇对比、与我研究方向的关联、 第几页提到 X、这篇论文的数据集怎么构造的、论文精读、 paper summary、summarize this paper、compare these papers、literature skim、extract methodology。 不做：论文检索、文献管理、论文翻译、扫描件 OCR、论文写作 / 降重。

绘制研究领域或技术主题的谱系脉络与历史演进图，可视化思想的演化路径，展示早期工作中的技术难题如何被后续研究逐步解决。当用户想了解某个主题的发展轨迹、某个模型或技术的"家族树"（family tree）、某条研究线索在多年间的演进路线、技术迭代脉络、论文/模型谱系，或询问"X 是如何一步步发展来的""X 解决了前人的什么问题""梳理 X 的发展历史"时触发。产出为嵌入 Mermaid 图表的 Markdown 文件（演进图 + 节点明细表 + 阶段叙事），可直接粘贴至 Mermaid 查看器或提交至代码仓库。

电商选品与品类机会分析技能。根据商家的预算、目标平台、货源优势，按"市场需求×竞争度×利润空间×季节性×复购率"五维打分，输出3-5个候选品类及切入建议。适用于新手不知道卖什么、老商家想拓新品类、想判断某个品值不值得进货的场景。

SEO Article Writer (Full Production Pipeline)

Scaffold a complete knowledge system. Detects platform, conducts conversation, derives configuration, generates everything. Validates against 15 kernel primitives. Triggers on "/setup", "/setup --advanced", "set up my knowledge system", "create my vault".

Apply plugin knowledge base updates to an existing generated system. Consults the Ars Contexta research graph for methodology improvements, proposes skill upgrades with research justification. Never auto-implements. Triggers on "/upgrade", "upgrade skills", "check for improvements", "update methodology".

This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.

Interact with Jira from the command line to create, list, view, edit, and transition issues, manage sprints and epics, and perform common Jira workflows. Use when the user asks about Jira tasks, tickets, issues, sprints, or needs to manage project work items.

Auto-create new skills from repetitive workflows. When you notice yourself doing the same multi-step process repeatedly, extract it into a reusable SKILL.md that any agent can use.

API design principles and decision-making. REST vs GraphQL vs tRPC selection, response formats, versioning, pagination.

Socratic questioning protocol + user communication. MANDATORY for complex requests, new features, or unclear requirements. Includes progress reporting and error handling.

Web application testing principles. E2E, Playwright, deep audit strategies.

Generative Engine Optimization for AI search engines (ChatGPT, Claude, Perplexity).

Internationalization and localization patterns. Detecting hardcoded strings, managing translations, locale files, RTL support.

Manage and compress conversation context in long sessions. Detect when context is growing large, summarize completed work phases, archive old findings while preserving key decisions. Prevents context degradation.

Game art principles. Visual style selection, asset pipeline, animation workflow.

Use to finalize PNAS display items — column-width sizing (~9 / 11.4 / 18 cm), minimum font sizes, show-the-data with n and defined error bars, scale bars, colorblind-safe palettes, legend structure, image integrity, and color-mode notes. Enforces figure rigor before submission.

Use when auditing an ACM CHI Papers submission for PCS readiness, covering the September full-paper deadline, the anonymized single-column template, word-count norms instead of page limits, subcommittee designation, supplementary and video upload, anonymization of external links, and CHI's desk-reject triage before upload.

Use to convert references to NEJM's Vancouver / ICMJE numbered style — numbered in order of appearance, limited author lists then et al., NLM journal abbreviations, and a capped reference count. Late-stage style pass.

Use when planning an ACM SoCC project timeline from venue fit through the two-round abstract+paper deadlines, the rebuttal, notification, camera-ready, and presentation, with backward-planning offsets for a cloud-systems measurement paper and honest handling of the two-rounds-per-year cadence and the no-cross-round-resubmission rule.

Refactor changed code to reduce duplication, complexity, and wasted work. Use for diffs, commit ranges, PRs, paths, staged changes, or working-tree changes.

Discovers and applies curated prompts from the prompts.chat collection to optimize AI interactions, prompt engineering, and workflow integration.

Complete PortSwigger deep-dive with exact payloads for every lab variant including zero-day techniques

Full ASVS 5.0 compliance assessment against a codebase and/or architecture diagrams. Reads all 346 controls from the companion CSV, performs targeted code analysis per control, and produces a complete matrix marked COMPLIANT / NON_COMPLIANT / NOT_RELEVANT — with per-control reasoning and evidence (code snippets, file:line references, diagram observations). Outputs a reviewed CSV matrix and a self-contained HTML evidence report.

Complete PortSwigger deep-dive with exact payloads for every lab variant including zero-day techniques

Utilize Windows Management Instrumentation (WMI) to execute malicious payloads, establish lateral movement, and execute commands stealthily across an Active Directory environment without dropping binaries to disk or relying on traditional Service Creation (PsExec) mechanics.

Race condition and TOCTOU testing for web apps. Use when testing one-time operations, concurrent HTTP abuse, rate-limit bypass, Turbo Intruder gates, HTTP/2 single-packet attacks, and CWE-362-style synchronization gaps.

Exploit Active Directory environments using Pass-the-Hash (PtH). This skill details how to authenticate to remote systems using the NTLM hash of a user's password instead of the plaintext password, enabling lateral movement and privilege escalation without cracking hashes.

Exploit misconfigurations in AWS Cognito, specifically focusing on unauthorized identity pool access, user pool self-registration issues, and privilege escalation via custom attributes to access broader AWS infrastructure.

HTTP request smuggling and desynchronization testing. Use when front proxies, CDNs, or load balancers disagree with the origin on message framing (Content-Length vs Transfer-Encoding), on HTTP/2→HTTP/1 translation, or when exploring client-side desync via browser fetch pipelines.

Terraform infrastructure-as-code workflow patterns: state and environments, module design, safe plan/apply, drift control, and CI guardrails

Base reference tool. Use when working with base in blockchain contexts.

Manage blog posts with drafts, scheduling, and SEO optimization. Use when creating articles, optimizing metadata, or scheduling publication dates.

core reference tool

Defi reference tool. Use when working with defi in blockchain contexts.

Manage Trello boards, lists, and cards via the Trello REST API.

Create, search, and manage Bear notes via grizzly CLI.

Admin and authority operations for the Pump protocol — set coin creator, update token incentives, set IDL authority, claim cashback, Mayhem mode, and BothPrograms cross-program admin instructions.

Generate or edit images via Gemini 3 Pro Image (Nano Banana Pro).

Transcribe audio via OpenAI Audio Transcriptions API (Whisper).

当需要进入高效学习、创作或问题解决的心流状态时，调用此模型作为环境与心理准备框架

Guia abrangente para Manim Community - framework Python para criar animações matemáticas e vídeos educacionais com controle programático

Use quando o usuário solicita narração em texto-para-fala, voiceovers de acessibilidade, prompts de áudio ou geração em lote via OpenAI Audio API; execute a CLI incluída (`scripts/text_to_speech.py`) com vozes integradas e requer `OPENAI_API_KEY` para chamadas diretas. Criação de vozes customizadas está fora do escopo.

Adicionar, visualizar ou remover domínios para serviços Railway. Use quando o usuário quer adicionar um domínio, gerar um domínio railway, verificar domínios atuais, obter a URL de um serviço ou remover um domínio.

Padrões da biblioteca de componentes Material-UI v7 incluindo estilização com sx prop, integração de tema, design responsivo e hooks específicos do MUI. Use ao trabalhar com componentes MUI, estilização com sx prop, personalização de tema ou utilitários MUI.

Framework de computação quântica para construir, simular, otimizar e executar circuitos quânticos. Use esta skill ao trabalhar com algoritmos quânticos, design de circuitos quânticos, simulação quântica (com ou sem ruído), execução em hardware quântico (Google, IonQ, AQT, Pasqal), otimização e compilação de circuitos, modelagem e caracterização de ruído, ou experimentos e benchmarking quântico (VQE, QAOA, QPE, randomized benchmarking).

Use essa skill para processar e analisar grandes conjuntos de dados tabulares (bilhões de linhas) que excedem a RAM disponível. Vaex excels em operações DataFrame out-of-core, avaliação lazy, agregações rápidas, visualização eficiente de big data e machine learning em datasets grandes. Aplique quando usuários precisarem trabalhar com arquivos CSV/HDF5/Arrow/Parquet grandes, realizar estatísticas rápidas em datasets massivos, criar visualizações de big data ou construir pipelines de ML que não cabem em memória.

Kit de ferramentas para arquivos genômicos. Leia/escreva alinhamentos SAM/BAM/CRAM, variantes VCF/BCF, sequências FASTA/FASTQ, extraia regiões, calcule cobertura, para pipelines de processamento de dados NGS.

Use esta competência ao trabalhar com matemática simbólica em Python. Esta competência deve ser usada para tarefas de computação simbólica incluindo resolução de equações algebraicamente, realização de operações de cálculo (derivadas, integrais, limites), manipulação de expressões algébricas, trabalho com matrizes simbolicamente, cálculos de física, problemas de teoria dos números, computações de geometria e geração de código executável a partir de expressões matemáticas. Aplique esta competência quando o usuário precisa de resultados simbólicos exatos em vez de aproximações numéricas, ou ao trabalhar com fórmulas matemáticas que contêm variáveis e parâmetros.

Analise arquivos FCS (Flow Cytometry Standard) v2.0-3.1. Extraia eventos como arrays NumPy, leia metadados/canais, converta para CSV/DataFrame, para pré-processamento de dados de citometria de fluxo.

Kit de ferramentas abrangente de computação quântica para construir, otimizar e executar circuitos quânticos. Use quando trabalhar com algoritmos quânticos, simulações ou hardware quântico, incluindo (1) Construção de circuitos quânticos com portas e medições, (2) Execução de algoritmos quânticos (VQE, QAOA, Grover), (3) Transpilar/otimizar circuitos para hardware, (4) Executar em IBM Quantum ou outros provedores, (5) Química quântica e ciência dos materiais, (6) Aprendizado de máquina quântico, (7) Visualizar circuitos e resultados, ou (8) Qualquer tarefa de desenvolvimento de computação quântica.

Especialista em backend reativo Convex: design de schema, funções TypeScript, assinaturas em tempo real, autenticação, armazenamento de arquivos, agendamento e deploy.

Filtros de química medicinal. Aplique regras de similaridade a fármacos (Lipinski, Veber), filtros PAINS, alertas estruturais, métricas de complexidade, para priorização de compostos e filtragem de bibliotecas.

Rastreie sites inteiros usando a API /crawl do Cloudflare Browser Rendering. Inicia jobs de rastreamento assíncronos, monitora a conclusão e salva os resultados como arquivos markdown. Útil para ingerir sites de documentação, bases de conhecimento ou qualquer conteúdo web no contexto do seu projeto. Requer variáveis de ambiente CLOUDFLARE_ACCOUNT_ID e CLOUDFLARE_API_TOKEN.

Build throwaway code that answers exactly one design question — a tiny terminal app to feel out logic/state models, or several radically different UI variants switchable on one route. Use when the user wants to "prototype", "mock up", "sketch", "try a few versions", "see what it'd look like", "feel out this state machine", or is stuck choosing between designs they can only judge by seeing or driving them. Also used by the whiteboard skill for napkin-type tickets.

Reconstruct a product flow, then audit how usable it is — whether it makes sense, where users get confused or stuck, whether the copy and instructions are clear, and where friction causes drop-off. Produces a prioritized, severity-weighted findings report. Works from three inputs: frontend code in a repo, a live running app driven via a browser, or a written flow / spec / screenshots. Use this skill whenever the user asks for a UX review, usability audit, "is this flow confusing", "does this signup/onboarding/checkout make sense", "why are people dropping off", "audit this form", "where's the friction", or hands over a flow and asks whether it's user-friendly. Trigger even on casual phrasings ("is my checkout any good?") — it imposes the structure ad-hoc UX opinions miss. Audits flow usability and friction; for visual polish and interface quality the impeccable audit/critique skills apply instead.

Plan a chunk of work too big for one agent session by putting it on a shared whiteboard — a map of investigation tickets on GitHub Issues — working them until nothing is left to decide, then snapshotting the board into a handoff artifact. Tickets needing nobody are worked back-to-back; the session stops when the human is the blocker. Use only when the user explicitly invokes whiteboard or asks to draw, work, run, or snapshot a whiteboard/map — not for ordinary planning requests.

Analyze a website page's UI for obvious oddities: the same information repeated multiple times (a phone number shown three times in the header, a '24/7' badge in two places, a bullet list restating the paragraph above it), duplicated or near-duplicate copy and sections, stray/orphaned text near CTAs, copy that doesn't match the business or section subject, placeholder text, overcrowded heroes, misaligned elements, and imagery that doesn't fit the business. Use this whenever the user asks to check, audit, review, sanity-check, QA, or 'look over' a web page, landing page, or UI — whether they give a live URL, a localhost/dev-server URL, a screenshot, or HTML/JSX/template files — even if they only say something casual like 'does this page look right?', 'find anything weird on this page', or 'anything off about my landing page?'.

Design the architecture for a multi-command skill system — the shared state schema, the command map, the router, and the boundary each command stops at — before any skill files get written. Use when the user wants to turn one skill into a system, combine several related or overlapping skills into a single coherent one, build a skill with subcommands (like `/thing audit`, `/thing polish`), design a suite, family, or collection of skills, or asks "how should I structure this skill set". Also trigger on "skill system", "skill architecture", "multi-command skill", "system of skills", "make these skills work together", "skill with slash commands", or complaints that their skills overlap, collide, duplicate each other, or never fire. Produces an architecture spec only — it never writes SKILL.md or reference files. Hand the finished spec to skill-creator to build from.

Retrieve task-relevant project and global memory without loading everything. Load when the user asks what we decided, recall prior context, find memory about a feature, explain past rationale, resume a task, or check deferred ideas.