Explorar skills

Rules and examples for adding Sentry metrics the right way. Covers how to name a counter, gauge, or duration metric; which tags are safe versus which will blow up your Sentry bill; how to track failures with a small fixed list of error types instead of raw exception strings; and how to add metrics around HTTP routes, external API calls, workflow steps, retry loops, and fallback paths without copy-

Fetch Bilibili (哔哩哔哩) video comments for UP主 self-analysis. Use when the user asks to collect, download, export, or analyze comments on a Bilibili video (BV号 / URL / UID). Produces JSONL + summary.json suitable for further Claude Code analysis (sentiment, keywords, audience trends). Read-only; does not post/edit/delete.

Use when working with Jira or Confluence from command line, including authentication, searching issues with JQL, bulk operations, sprint reports, or creating/updating work items using acli

为 GitHub 项目生成符合最佳实践的专业 README.md，包含 SEO/GEO 优化标题、16:9 信息图三件套（banner/features/workflow）、结构化作者信息、MIT 许可证、GitHub description 和 Topics 推荐。当用户说「帮我写 README」「生成 README」「README 最佳实践」「README 信息图」「GitHub 项目首页」「readme maker」「readme 生成」「写一个专业的 README」「项目文档」「开源说明」时必须使用此 skill。也适用于 README 已存在但需要重构、补充信息图、添加作者信息、优化排版的场景。

Personal engineering rules and reasoning patterns the user has curated over time. Consult BEFORE writing, modifying, debugging, or designing any technical feature — rules often cover domain-specific gotchas (e.g., mobile audio timers need background mode) that generic answers miss. Also triggers on `/learn-this`, "learn this", "remember this", or `/new-rule` to capture new rules, and on detected c

Production-grade LLM integration patterns for Ruby on Rails applications. USE THIS SKILL when the user needs to: integrate OpenAI, Claude, or any LLM API into a Rails app; use ruby_llm gem or langchain-rb gem in Rails; build AI-powered features in Rails; manage prompts as templates; route between LLM models for cost optimization; run LLM calls as background jobs; set up Braintrust evals; use LiteL

Find academic papers across 5 sources (OpenAlex / Semantic Scholar / CrossRef / PubMed / arXiv) with adjustable depth — Quick scan (5 min) to Audit prep (3 hr). Use when the user wants to find papers, run a literature search, gather references, or scope a research topic. Triggers on: search verbs ('find papers', 'literature search', 'papers about X'), review types ('scoping review', 'systematic re

Closes the agentic coding loop for web animation. Lets Claude scroll the user's page while recording it (Playwright), extract frames, build a labelled contact sheet, and reason about timing/easing/trajectory frame-by-frame — plus motion design principles (Disney's 12, adapted for web, GSAP, CSS, scroll input) to name and fix what's wrong. Use whenever the user wants to add, improve, debug, or revi

Agende e publique conteúdo (texto, imagens, carrosséis e vídeos) em redes sociais como Facebook, Instagram, LinkedIn, TikTok, YouTube, Pinterest e Telegram, utilizando a API POST da Devad.io. Esta skill permite postar conteúdo programaticamente, sem a necessidade de um painel de controle de mídia social.

Autonomous Goal-directed Iteration. Apply Karpathy's autoresearch principles to ANY task. Loops autonomously — modify, verify, keep/discard, repeat. Supports optional loop count via Claude Code's /loop command. Invoking /autoresearch <free-form goal> builds a real-data benchmark harness, captures a baseline, and iterates with a regression gate until the goal is hit.

Use this when the user wants to generate, normalize, verify, deduplicate, or export training datasets for Codex, Antigravity, or Claude Code from topics, URLs, reference material, web research, or existing JSONL/CSV files. Supports SFT and DPO workflows, custom export schemas, and deterministic local pipeline scripts.

Esta habilidade de colaboração em pensamento de produto auxilia usuários a esclarecer ideias, validar necessidades, entender tarefas, convergir para um MVP, diagnosticar problemas e planejar verificações. Ela opera identificando tipos de tarefa e incertezas, entrando então em um modo (diagnóstico, fundador ou construtor) para aplicar regras e ações específicas, fornecendo observações, julgamentos, sinais de alerta, ações e planos de verificação.