conteudo-instagram-skill
Skill para Claude Code: do concorrente ao post agendado. Base de conhecimento do nicho, branding, carrosséis de conteúdo e produto, posts estáticos em PNG 1080x1350 — para qualquer nicho do Instagram.
The most recently added listings, newest first. To search them, choose “Newest on findsafeskills” in search.
Skill para Claude Code: do concorrente ao post agendado. Base de conhecimento do nicho, branding, carrosséis de conteúdo e produto, posts estáticos em PNG 1080x1350 — para qualquer nicho do Instagram.
"AI video editor for talking-head reels. Turns a raw phone recording of someone talking to camera (retakes, pauses, false starts) into a finished vertical 1080x1920 reel for Instagram, TikTok or Shorts: the best take of each sentence, pause trimming, word-timed captions that never cover the face, motion scenes on the spoken word, sound effects mixed against the voice and checked B-roll. Use when the user hands over a talking-head video and says \"edit this\", \"make a reel\", \"make it look like this video\", \"add captions/sound effects\", \"pick the best takes\", \"remove the pauses\", asks for a re-cut, or names a style. Named styles (recipes/, docs/PLAYBOOK.md): Clean OS, Seoul list, RESULT, Face RESULT, stan style, vector explainer, POV montage, Whiteboard, Kinetic doc, Step Deck, Product page, Tactile paper, Caption News, News Package, Telestrator, Halftone Dossier, Swiss Grid, Data Story, Hologram HUD; plus seven presets (clean, editorial, meme, cinematic, launch, kinetic, pop). Not for landscape screen recordings."
说说你梦到了什么,它会翻古书讲讲古人怎么看这个梦,再用心理学聊聊它跟你最近的心事有什么关系。A skill for reading dreams: tell it your dream, and it shares what old Chinese dream books say, then helps you think about what it might mean for you.
Debloat one rogue Windows app: map launch chain, gates and self-repair, then durably block its ads, AI add-ons, telemetry and updaters. Not Windows-wide debloat or uninstall.
Find active accounts for leavers, orphaned and dormant accounts from an HR export and an app user list. Browser page, CLI and Claude skill.
Explain any codebase with hand-drawn parables that mean something: one drawing per real component, counts from real data, labels in any language (RTL too), 9:16 scenes and a captioned reel. A skill for Claude Code and coding agents.
Skill de Claude Code para estudiar para pruebas: resumen con sigla, flashcards y ruta de estudio con videos
Set up a second Claude account (e.g. personal + work) side by side in the Claude desktop app on macOS, each with its own login, chats, Code-tab sessions and Claude Code config, using a launcher app with its own badged icon. Use this whenever someone on a Mac wants to use two Claude accounts at once, switch between a personal and a company/work/client Claude account without logging out, run a second Claude desktop instance, or asks how to keep work and private Claude separate — even if they don't say "launcher" or "second instance". Also use it to troubleshoot or remove such a setup (wrong account after login, missing launcher icon, "already running" notice).
Cupraxis — Claude Code skill + source-tagged knowledge base for PCB design: floorplan, power, RF, EMC, components, DFM, KiCad 10 toolchain and vendor-agnostic fabrication (JLCPCB, PCBWay, Aisler, Eurocircuits…)
审查中文学术文本的 AIGC 检测争议、空泛论证与作者证据。用户提到降 AIGC、知网、学习通、PaperPass、维普检测结果或原创误判复核时使用;可提供本地开源检测器诊断和存证清单,但不预测商业平台分数或承诺达标。
Методика и каркас сайта на WordPress: блочная тема без сборки (Gutenberg/FSE) + плагин данных, удобный редактор, читаемый код с линтерами и пределами размера, SEO с первого дня (Rank Math кодом, JSON-LD, llms.txt, SEO-гейт в CI), безопасность, деплой по SSH через GitHub Actions с откатом. Используй, когда нужно: сделать сайт на WordPress по макету или прототипу; перенести старый сайт на новую тему с 301; настроить CI/CD на виртуальный хостинг; сделать блоки, которые удобно править в редакторе; настроить SEO/GEO, скорость, Polylang, Contact Form 7 + Telegram, аналитику с согласием на cookie; провести аудит такого сайта; навести порядок в коде темы или плагина. Triggers: блочная тема, block theme, FSE, Gutenberg, theme.json, block.json, render.php, сайт на WordPress, деплой WordPress, wp-cli импорт, перенос сайта, Rank Math, kalaganov5.
A Claude skill: start sharp, decide with numbers (STOP, PIVOT or CONTINUE).
Build a web-agency sales pitch deck from a prospect's website URL, or from their Facebook page or listing when they have no website (optionally with notes or voice notes about the prospect). Researches the live site, captures real screenshots, scores it, and publishes a slide deck that shows where the site falls short and what the new site would do instead. Use when the user sends a business website and asks for a pitch deck, sales deck, audit deck or teardown deck.
Claude skill that turns a start and an end point into an interactive 3D satellite trip map: real route, stops with photos, day plan, fuel, food and stays.
A Claude skill: a real pre-launch gate for sites and apps (security, legal FR/EU, Lighthouse).
Turns lecture recordings plus course slides into exam-ready study material, in two modes. Mode A: one Zoom share link with Zoom's official cloud transcript + that lecture's PDF/PPTX -> a bilingual knowledge-point Markdown (English verbatim + Chinese gloss) with every original slide rendered and embedded at full size, plus a self-contained PDF. If the recording has no official transcript, the skill downloads the audio only and stops; it never transcribes itself. Mode B: two or more finished lecture .md files + exam outline / past papers -> an 11-part TEXT-ONLY exam review handbook in Markdown and PDF. Triggers: 把录音整理成知识点, 总结这节课的知识点, 生成 lecture 讲义, 生成复习大纲, 期末复习手册, 根据所有 lecture 整理考点, 转成 pdf, or a Zoom recording link for a course. Review mode needs at least two lectures.
拿一份现成简历和一个目标岗位 JD,把简历改成贴合这个岗位的版本:拆 JD → 逐条对照给匹配度 → 用 JD 的原词重写经历 → 补岗位基本功 → 调语言风格 → 对话里确认后导出一页 PDF。给自己改、给朋友或线索改都用这一套。凡是用户丢来 JD(截图、链接、文字)加简历,说「看看这个 JD / 匹配度怎么样 / 帮他改简历 / 针对这个岗位改一下 / 把经历改成和 JD 贴合的语言 / 语言不够 B2B、不够产品 / 生成简历 PDF」,或者只丢来 JD 问「能不能投」时,务必使用本 skill,即使用户没说「定制」「tailor」。不用于:每天搜岗位并批量投递(job-hunt-autopilot)、给线索出简历诊断图组和话术(lead-resume-diagnosis)、从零访谈写第一份简历。
Free Claude skill: Amazon KDP checklist for Kindle eBooks & paperbacks. Format, print specs, metadata, AI disclosure, avoid rejection.
C 盘空间体检与清理。扫描占用点,清理可再生成的缓存、日志、已装软件的旧安装包。 Use when the user says "C盘没空间了"、"帮我清理磁盘"、"磁盘满了"、"看看什么能删"、 "清理C盘"、"磁盘清理"、"D盘也不够了"、"空间不够用了",或报告磁盘空间不足。 也用于排查空间黑洞(-Deep 定位已知类别之外的大头)。 Do NOT use for: 删除单个文件/文件夹(直接用文件工具)、卸载软件(用系统卸载程序)、 整理用户文档或照片(那是用户数据,不是缓存)、磁盘分区或格式化。
Build a polished static presentation from a single data file — design first, then website. Use when someone says "/deck-kit", "make a presentation/slide deck/pitch", "build a deck for X", or wants a talk they can run in the browser and deploy to their own server. Produces slides from deck/slides.json with an overview, phone reading mode, PDF export and a second presenter-notes window.
A Claude skill: merge and improve skills instead of multiplying them.
A Claude skill: plain, precise A4 PDFs from Markdown, no half-empty pages.
从模糊需求出发,定位 GitHub 上整块匹配的成熟模块(发现与核验走 github 连接器的 MCP 工具),深读代码提取事实并裁决优缺点,反推带验证信号的分步实现路径。触发词:帮我实现某个功能/模块、有没有开源项目可以参考、别人是怎么做这个模块的、需求很模糊想落地成方案。
从真实工作项目源码、Git 提交、任务系统、GitHub 与历史简历中挖掘可验证的个人技术贡献,通过单问题访谈确认后,生成中文 Markdown/Word/PDF 程序员简历。当用户要求根据项目源码、仓库、提交历史、禅道/Jira 整理简历、优化简历、匹配 JD 或提炼项目亮点时使用。禁止编造经历、技术栈、职责或量化成果。
An SDLC toolkit for AI coding agents: spec → plan → build → verify → review → ship.
"引证溯源(yinzheng-suyuan):本地整理参考文献并逐条追溯到源头、核验真实性、归档全文,仅供个人学习研究使用。两个取候选入口——①从 GradPaper(gradpaper.com/dissertation 智能写作)按题名+学历+学科专业+学校取它推荐的 18 条;②直接在知网(kns.cnki.net)按题名/主题自检索,按相关性、年份、被引、文献类型、排除词筛选。两条来源产出同格式 refs.json,之后做真实性核验、从开放渠道取全文、并在每份全文正文前插入一页「审核页」(题名/出处/核验结论/GB-T 7714 与 APA 引用串/文章链接/文件指纹)后存到本地。当用户说 gradpaper 推荐文献、抓推荐引文、按题目取参考文献、从 gradpaper 导文献、去知网搜参考文献、知网检索筛选文献、审核参考文献是否真实、核对引用出处、下载这批文献全文、给文献加审核页/引用页/封面页、参考文献溯源 时使用。也用于只跑其中一段:仅抓取(-DryRun)、仅知网筛选、仅核验、仅下载全文、仅加审核页。使用前必读包内 DISCLAIMER.md 与 USAGE-POLICY.md;脚本不含也不接受绕过鉴权、破解、验证码绕过、并发爬取等用途。"
Measure what changed after a website's SEO or AI search fixes go live. Reuse the original queries, target URLs and AI prompts, verify the live change, then compare matched Search Console windows or matched citation samples. Reports which fixes shipped, which outcomes changed, and what remains unknown. Use when asked whether a deployed search visibility fix worked, whether a page gained impressions, or whether AI answers cite a page more often. Not for the initial diagnosis (use diagnose-search-visibility), implemented fixes on individual skill pages (use fix-skill-visibility), or requesting indexing (use get-indexed).
Use when writing, editing, or auditing customer-facing UI copy, landing pages, pricing cards, onboarding, settings, forms, empty states, toasts, loaders, error messages, changelogs, SEO metadata, or llms.txt, especially when prompts mention internal release status, review queues, storage or AI architecture, billing blockers, or bare UI slots that tempt filler prose and AI marketing cliches.
pstack (cursor/plugins, MIT, Lauren Tan) for Hermes: the full upstream skill set, with a Cursor-to-Hermes tool adapter bundled in each skill.
AI-agent skill that audits ML code and notebooks for data leakage, test-set misuse, wrong metrics and training bugs, with a 0-100 score and fixes. Showcase and real-world results.
Agent Skills for customer acquisition: find verified leads, draft personalized outreach, and boost brand visibility in search and AI answers.
Method template for LLM-maintained wikis whose quotes, dates and figures are mechanically checked against linked sources. Built on an Agent Skill for Claude Code.
Shows which Claude Code skills your agent has quietly stopped seeing.
给一句需求,从 GitHub 上挑出该装哪个 agent skill。Find and vet agent skills on GitHub for a specific need.
Auditoria de segurança de funcionalidades de IA pelo OWASP Gen AI / LLM Top 10 (2025). Use ao revisar chats e assistentes com LLM, tool calling/agency, system prompts, prompt injection direta e indireta, renderização da saída do modelo (XSS), limites de custo/tokens, RAG/embeddings e supply chain de SDKs de IA, ou antes de publicar features de IA. Para AppSec clássico use owasp-security-reviewer.
Write the contract before the code, get the docs back from the code. spec → SPEC.md, prd → 左图右注 interactive PRD. Agent skill for Claude Code, Cursor, Codex and Windsurf.
Agent Skill: generate ATS-friendly CVs tailored to a job posting from a living career profile, with claim verification and no fabrication.
24 portable Markdown skills for AI agents, covering research, data analysis, documents, coding and review.
AI agent skill to audit & polish web UI against 20 motion principles & UX laws. For Claude Code, Gemini CLI, Cursor, and skills.sh.
Agent Skills for Claude Code and Codex, native Claude Code subagents, and portable role briefs. Historical skill evaluations on Claude models.
Apply and verify evidence-backed fixes for individual agent skill pages missing from Google, using a check-skill-visibility report or equivalent URL-level findings. Use when asked to fix a completed skill visibility diagnosis, repair identified canonical or discovery gaps, improve mismatched task copy, or correct stale listings for the affected skills. Produces a change and verification record tied to each finding. Not for a first diagnosis (use check-skill-visibility), a sitewide SEO sweep (use seo-sweep), or submission alone (use get-indexed).
Import AI skills written for other agents (Claude Code, Cursor, Codex, plain Markdown) into DeepSeek Harness, and publish them as native DSH skills.
Read a Zi Wei Dou Shu (紫微斗數) chart in the traditional order — Life palace, then the
Verificação funcional e QA cético de mudanças de código. Use depois de implementar uma feature, antes de fazer merge de um PR, ou quando for preciso confirmar que um trabalho "pronto" realmente funciona — descobre e executa lint, typecheck, testes e build do projeto, desenha casos de teste (happy path, borda, autorização, erro) e reporta o que passou, falhou ou não pôde ser verificado.
Pin agent skills installed with npx skills to commit SHAs, and gate every upstream change behind isolated LLM reviewers
Guardrails for AI agents that use iOS simulators, so parallel agents don't melt the Mac. Use before any session boots, builds to, runs on, taps through or screenshots an iOS simulator (Xcode MCP, xcodebuild, simctl, XcodeBuildMCP, ios-simulator-mcp, Maestro), and when the user asks why the Mac is hot, slow or out of memory.
AI 产品经理简历 Skill 介绍与 ¥9.9 微信/支付宝购买入口。
Agent Skill for Claude: plan a sauna session and size a heater from published rules.
AI 能力画像:让你常用的任何 AI 根据真实使用记录,生成 30 项能力评估报告(雷达图 + 证据 + 改进建议)。Claude Code、Codex 等装 skill 即用;DeepSeek、千问、WorkBuddy 等网页版发文档即可。| AI Profile: any AI you use assesses 30 AI-usage skills from your real history. Install as a skill, or send the doc to any chat AI.