Recently Added
Fresh skills just added to the collection. Be the first to try them out!
ce-commit-push-pr
by EveryInc
Commit, push, and open a PR. Use when asked to ship/open a PR, or for PR-description-only flows like writing, rewriting, or describing a PR body.
server-doctor
by evgyur
Use when auditing or repairing Linux or macOS hosts that run OpenClaw, Telegram bots, or related automation. Start with host/runtime/access mapping, then use focused references for incidents, updates, Task Flow, and safe public documentation.
meteo-cnews
by gregorylanglet59264-byte
Génère et met à jour automatiquement les cartes météo HD (Matin, Après-midi, Soirée) nationales (J+7) ou régionales (J+2) et les rapports CSV correspondants pour CNews à partir de l'API de Météo-France.
whatsapp-claude
by tomerab1
Watch ONE WhatsApp group and reply when a member tags @claude (or @קלוד). Replies are generated by a hardened, sandboxed headless claude -p (WebSearch only; no fs/shell/secrets/MCP) using recent chat context, and posted from the user's own paired account, prefixed 🤖. Sibling of whatsapp-kg (own linked device, own data dir, own always-on pm2 daemon). Use to add a summonable group assistant to a group.
design-mvr-brief
by oavvis
Run Design MVR minimum viable research for brand, product, campaign, content, website, UI, or personal-brand design work. Use when a user wants to collect brand-owner information, clarify brand purpose, mission, how and why the brand creates value, perform brand strategy analysis, audit competitors or visual references, define anti-references, translate strategy into visual motifs and searchable reference keywords, or produce a creative brief with all context needed before visual design begins.
posthive
by AstaBlackClove
Schedule social media posts to (Bluesky, Threads, Instagram, LinkedIn, Mastodon, YouTube, Facebook Pages, Pinterest, Telegram, X/Twitter, Nostr, Discord, Tumblr and more) via the Posthive CLI. Use when the user asks to create, schedule, list, update, or delete social media posts, or manage their posting queue.
expert-ai-systems-engineer
by Rene-Kuhm
"Trigger only when the user asks to rewrite, translate and strengthen, professionalize, or convert rough software/AI requirements into a copy-ready English engineering prompt. Exclude direct execution and literal/verbatim translation-only requests."
explaindraw
by NLR-2007
Create clear visual explanations from natural-language requests as mascot illustrations, editable PowerPoint diagrams, or hybrid illustration-plus-diagram slides. Use when a user asks to draw, illustrate, visualize, make a flowchart, architecture diagram, timeline, roadmap, conceptual analogy, educational visual, or hybrid explanatory slide, especially when they provide this repository URL and ask the AI to use it as a skill.
boss-zhipin-jd-scraper
by HeyClioo
Scrape full job descriptions (responsibilities & requirements) from BOSS Zhipin / zhipin.com. Use when the user wants to scrape or batch-collect BOSS Zhipin job postings, JDs, or recruitment data by filter. Needs a browser-automation tool that drives your real, logged-in browser (Claude Code's Chrome extension, or an equivalent in Codex / WorkBuddy) — NOT headless — to bypass anti-bot verification; strips watermark noise, dedupes by encryptJobId, exports Markdown. 抓取 BOSS直聘职位的岗位职责/任职要求全文；需用能驱动真实已登录浏览器的工具（Claude Code / Codex / WorkBuddy 等），去重去水印导出 Markdown。
ai-stock-board
by xiaomin4576-ui
美股 AI 科技股早报。以二级市场交易员/buy-side 视角,长期(6-12月)跟踪「美股 AI 产业链上下游核心十票 + QQQ」(英伟达、博通、台积电、超微、美光、迈威尔、Arista、Credo、相干、Vertiv + QQQ),每支给两个核心指标:①买入建议(买入/观望/回避+建议买入价)②6-12月目标价+预期收益。内置"预测台账→自动复盘→校准"闭环,让买入价/目标价越用越准。引擎是 Claude 本人(不用 DeepSeek),数据用 yfinance 真实行情。当用户说"跑美股AI早报/AI股票早报/科技股早报/英伟达们今天怎么看/ai-stock-board"或需要定时盯这组票时使用(本地出看板)。另:说"刷新云端美股看板/把美股早报推到线上/跑美股AI早报云端版/更新线上美股看板"时,走【云端手动刷新】——Claude 亲自研判一期并 push 到云端仓库,触发线上 Pages 重建+飞书推送。
wikime
by Phluxxed
Use when user runs /wikime or asks to scaffold a new wiki, knowledge base, or LLM-maintained document store in an empty or fresh directory
performance-prompter
by Gleam-studio-project
读取剧本与人设，为 Seedance 2.0 生成文戏/情感戏表演提示词。当用户要生成表演提示词、提升 AI 视频人物表演质量、把剧本场次转成 Seedance 提示词时使用。主打欧美市场海外短剧（英文对白），兼容国内题材；专攻文戏/情感戏，不适用于动作戏。
literature-workflow
by hwang847
"Use for local literature workspaces: organize and read papers, map copied titles to local PDFs, manage paper and survey sources, prepare scoped survey work packs, create concise HTML notes, and hand off selected research artifacts to Feishu/Lark."
tsao-sci-researcher
by SUNHAOJUN22
Evidence-first, full-lifecycle scientific research skill for research-question formulation, deep literature research, systematic review, experimental design, data quality and statistics, publication-grade scientific figures, manuscripts, technical reports, peer review, project governance, patents, laboratory workflows, biomedical research and research integrity. Use when a user asks to design, investigate, analyze, visualize, write, review, audit, manage or communicate scientific research. Delegate real multiscale simulations to TsaoSciComputation through the structured computation-handoff workflow.
扫地僧
by J-levee
扫地僧 — 科研方法论谋士（研究设计/方法选型/批判思维/卡点破局）。藏经阁·易筋出品。
godot-engine-4-7
by sxdy4591
Use when writing GDScript or C# code for Godot Engine 4.x projects, creating game scenes, configuring physics/navigation/shaders/UI/audio/animation/networking/XR, debugging Godot projects, or answering any Godot Engine development question. Covers complete GDScript and C# syntax, all 1083+ API classes, node system, 2D/3D rendering, physics, shaders, UI, animation, audio, navigation, input, networking, file I/O, exporting, performance optimization, XR, and engine architecture.
video-understanding
by stevederico
Turn any video (local file or X post) into a complete AI understanding: extract timestamped frames + a transcript, then correlate what's on screen with what's said into an understanding.md. Runs fully local (whisper.cpp + ffmpeg, no keys) or install-free via BYOK (xAI STT + Mux). Use when the user says "video understand this", "understand this X post", "analyze this video", or runs /video-understanding.
agentic-ui-qa
by HomenShum
Universal QA + dogfooding protocol for agentic application UIs (worked profiles included for NodeBench AI, NodeRoom Live, NodeSlide/parity-studio; template for any app). Use when asked to QA, dogfood, verify, regression-test, or "revamp the agent UI" of ANY app — it drives the UI end to end as personas, verifies every claim with artifacts, scores the app against the Agentic UI Bar, and runs the bounded fix-revamp loop. Agent- and model-agnostic — a literal floor any small model can execute, an additive ceiling that scales with stronger models.
vibe-scratch-workflow
by mizchi
vibe new から vibe shell-stdin --restore で scratch-db を積み上げ、finalize/normalize/apply まで安全に回すための実践手順。
ce-proof
by EveryInc
Publish, read, comment on, or edit markdown in Proof. Use for Proof links, sharing specs/plans/drafts, or publish handoffs from planning workflows; avoid proofread, math, evidence, or proof-of-concept meanings.
c2py23
by jonwright
Generate zero-copy C99 Python extensions from declarative YAML interface files. Use when wrapping C functions for Python, creating high-performance numerical extensions, binding SIMD-optimized C code, or when the user mentions c2py23, C extensions, or buffer-protocol wrapping. Supports Python 2.7-3.15, GIL release, free-threading, and automatic buffer shape/type dispatch.
agent-factory
by n4rly-boop
Spawn and talk to long-lived SECONDARY Claude agents from this session — real peer agents in their own tmux sessions, not ephemeral subagents/Tasks. Use this whenever the user wants to launch/spawn/open another agent or "a second Claude", drive or chat with a peer/worker agent, run agents that talk to each other, revive a killed agent, or list/clean up previously spawned agents and their session logs. ALSO use it to stand up a whole TEAM/FLEET of agents with roles and a hierarchy (orchestrator + workers, "agents that each own a role", parallel multi-agent experiments, ablations) — that is line.sh with a blueprint.yml. Trigger even if the user just says "spawn an agent", "open another claude", "talk to it", "kill that agent", "set up a team of agents", or "clean up the agent logs" without naming this skill.
anlin-writing
by github-yjc
"Generate, review, or evaluate Anlin/日寄-style anonymous blind-evaluation prose against the Anlin corpus. Use ONLY when the user explicitly mentions anlin-writing, anlinwriting, Anlin, 日寄, Anlin-style, 像Anlin那样写, 模拟日寄, 盲评, or asks for Anlin corpus evaluation. In any workspace containing .anlin-clean-eval-mode, check that marker before drafting and use clean_run_checker.py, not the normal checker."
batch-translate
by xiaoxinblast
批量翻译工作流。支持 mqxliff/docx/xlsx/txt → 分批翻译 → 逐批校对 → 写回，全自动循环。 触发条件：用户说"开始批量翻译""批量翻译这个文件""继续翻译"等。