Recently Added

glyph

by sumitttt4

A brand and design skill for AI-built products. Use it to create brand identity, UI direction, typography, color systems, landing pages, dashboards, product taste, anti-generic design rules, and Vibe Coding prompts.

windows-ad-pentesting

by ErenYeager29

Windows AD pentesting beginner to master. Trigger on: Kerberoasting, AS-REP Roasting, Golden/Silver/Diamond/Sapphire Tickets, Pass-the-Ticket, Overpass-the-Hash, noPac, unconstrained/constrained/RBCD delegation, KrbRelay, Bronze Bit, Timeroasting, BloodHound/SharpHound, lateral movement (PsExec, WMI, WinRM, DCOM, PtH), ACL abuse (GenericAll, WriteDACL, GPO, AdminSDHolder, shadow credentials, DNSAdmin), DCSync, credential dumping (Mimikatz, secretsdump, LSASS, DPAPI, DonPAPI, domain backup key), NTLM relay (Responder, ntlmrelayx, PetitPotam), AD CS (ESC1-16, Certifried, Certipy, PKINIT), MSSQL (mssqlclient, xp_cmdshell, linked servers), gMSA, Windows LAPS, ADIDNS/wpad, SCCM/MECM/SharpSCCM, WSUS, persistence, forest/trust attacks, AMSI/ETW/process injection evasion. Vague triggers: "pwn AD", "get DA", "domain admin", "krbtgt", "ntds.dit", "AD lab". All topics: concept, commands, OPSEC, detection.

ai-dev-workflow

by Melody-YK

Orchestrate a lightweight, artifact-driven AI development workflow from PRD/intake through requirements analysis, product/engineering review, static prototype generation, implementation planning, build, and verification. Use when the user wants to run requirements-analyst, gstack, and superpowers together in a controlled workflow, create .ai-workflow artifacts, generate a requirements-driven prototype, or test an AI development process on a PRD.

gf-marketing-planner-v2

by martinariasf

Read or update the per-client data files that drive the Viktor Marketing Operating Dashboard. Use when adding posts, scheduling content, planning a quarter, updating client context, recording learnings, or syncing performance metrics. The agent writes JSON files in marketing-planner/clients/<slug>/; the React dashboard reads them live. TRIGGERS — any of: "draft a post", "add post for [client]", "approve <post-id>", "update brief for [client]", "plan Q[N] for [client]", "sync metrics from Postiz", "log a learning". DO NOT trigger for: general copywriting unrelated to a tracked client, one-off posts not going into the dashboard.

cde-query

by yygtmp

鏌ヨ鍥藉鑽搧鐩戠潱绠＄悊灞€鑽搧瀹¤瘎涓績锛圕DE锛夊師杈呭寘鐧昏淇℃伅锛岃嚜鍔ㄧ炕椤?鑷獙璇侊紝杈撳嚭 Excel銆傛敮鎸佹悳绱㈠師鏂欒嵂銆佽嵂鐢ㄨ緟鏂欍€佽嵂鍖呮潗鐨勭櫥璁板彿銆佷紒涓氬悕绉般€佸璇勭姸鎬併€?license: MIT

plain-language

by JunyoungOh

Use when a non-developer needs Claude to communicate in plain, non-technical language. This skill changes HOW Claude talks in the conversation — it rewrites explanations, plans, and progress updates into jargon-light Korean with everyday analogies, keeping technical terms but glossing them on first use. It also runs a concrete one-time setup: it writes a persistent switch into CLAUDE.md so plain-language stays on across future sessions, and removes that switch when the user wants to turn it off. Because of this setup machinery, invoke the skill rather than just answering casually — especially when ongoing or "from now on / 앞으로" behavior is implied. Trigger whenever the user says they are a non-developer / 비개발자 / 초보자 / 코딩을 모른다, asks Claude to explain or speak 쉽게 / 쉬운 말로 / 풀어서 / 비유를 들어서, says they can't follow the technical jargon, wants the conversation made understandable for a non-developer who will read it (a parent, a friend, a teammate), or invokes /plain-language — even without naming the skill. Do NOT trigger for requests to simplify or refactor code, rename variables, tidy code comments or commit messages, rewrite a README or API docs, declutter a UI, or translate text into simpler language — those edit an artifact, not the way Claude converses.

suanfish-design-system

by SuanFishXYY

Suanfish Design System — A unified multi-agent design language system with 14 specialized agents across 6 tiers and a REJECT mechanism for opinionated AI-driven UI generation. Governs both first-impression onboarding moments (welcome, version updates) and steady-state interfaces (three-pane layouts, modals, wizards, data viz). Activate when designing, generating, auditing, or refactoring any UI surface. 算鱼工作室的统一多智能体设计语言体系，14 位专精 agent 协同工作，分布在 6 个 tier，搭载 6 条 REJECT 硬规则。

ros2-engineering-skills

by dbwls99706

TRIGGER when the user: writes or reviews ROS 2 nodes (rclcpp/rclpy), creates packages (colcon/ament), edits launch files (.launch.py), configures QoS or DDS, writes URDF/xacro, implements ros2_control hardware interfaces or controllers, sets up Nav2/MoveIt 2 pipelines, processes sensor data (camera/LiDAR/PCL), works with Gazebo/Isaac Sim, configures SROS2 security, develops micro-ROS firmware, manages multi-robot fleets (Open-RMF), debugs with ros2 doctor/rosbag2, deploys via Docker/cross-compilation, or migrates from ROS 1. DO NOT TRIGGER for general C++/Python questions unrelated to ROS 2, non-robotics middleware, or web/mobile development tasks.

AplicacionMovile

by TheKidin

verification-convergence

by moonhwilee

"High-intensity /verify or /converge protocol for specialist subagent review, selective fixes, and evidence-based convergence across code, plans, docs, strategy, business, trading, and operations."

ads-bib

by YifZhou

Use this skill for ANY request involving astronomical references, citations, BibTeX entries, or \cite{} keys — even if the paper seems well-known enough to recall from memory. Triggers include: "find the bib entry for...", "get me a reference to...", "add citations for...", "draft this with proper ADS references", "look up the paper by X about Y", or any request to include \cite{} keys in LaTeX. Also trigger when drafting or editing a scientific manuscript and references are needed — even if the user does not explicitly say "ADS" or "BibTeX". NEVER fabricate bibcodes or BibTeX fields, even for famous papers. NEVER use web search, memory, or any other source for bibliography data — ads_search and ads_bibtex are the ONLY permitted methods for all bibliography tasks, including LaTeX documents, papers, and proposals.

ultimate-ppt-master

by kdnsna

终极融合PPT大师 / Ultimate Fusion PPT Master: AI-driven presentation generation system. Converts source documents (PDF/DOCX/XLSX/URL/Markdown) into either editable PPTX decks or editorial/Swiss magazine web slide decks. Use automatically when the user asks to create or transform content into a presentation, slide deck, PowerPoint, PPT, or PPTX, including phrases such as "create PPT", "make presentation", "make a deck", "build slides", "turn this into slides", "生成PPT", "做PPT", "做一个ppt", "做一个 PPT", "做个PPT", "做个 PPT", "制作演示文稿", "生成演示文稿", "把这个做成PPT", "把这份材料做成PPT", "杂志风PPT", "网页PPT", "horizontal swipe deck", "editorial magazine presentation", or mentions "终极融合PPT大师", "ultimate-ppt-master", "deckweaver", or "ppt-master".

math-engine

by Arctic-Antarctic

确定性数学运算引擎。当用户需要精确数值计算、表达式化简/展开/因式分解、 符号微分与积分、或解方程时调用。把表达式交给本工具计算，不要心算或用语言模拟计算。 Deterministic math engine. Invoke whenever the user needs exact arithmetic, expression simplify/expand/factor, symbolic differentiation or integration, or equation solving. Always route the expression to this tool instead of computing it in your head.

forgeflow

by gimso2x

Artifact-first delivery workflow for AI coding agents

sms

by sunjinpak

Send SMS in bulk from this Mac via iPhone+Mac SMS forwarding, and (optionally) scan replies in the local Messages database. Use when the user wants to send a personalized text message to one or more recipients ("send a text to ...", "remind everyone in this list", "text the team about ..."). Free, no external API.

Keyword Tier 3 — State-Specific Programmatic SEO

by kolegadev

```

Keyword Tier 2 — Problem-Aware

by kolegadev

```

Keyword Tier 1 — Core Transactional

by kolegadev

First 50 keyword validations need human QA

Email Copywriter

by kolegadev

GITHUB_TOKEN — Auto-provided for Actions

Social Content Creator

by kolegadev

GITHUB_TOKEN — Auto-provided for Actions

Geo Content Generator

by kolegadev

GITHUB_TOKEN — Auto-provided for Actions

aaq

by OODALabs

"AI Acceleration Quotient (AAQ) is a methodology developed by Bob Gourley of OODA.com for scoring and ranking companies on AI maturity across machine learning, generative AI, and agentic AI. Use this skill whenever the user asks to score, rank, benchmark, or compare companies on AI capability, AI maturity, AI readiness, or digital transformation. Also trigger when the user mentions 'AAQ,' 'AI Acceleration Quotient,' 'AI scorecard,' 'AI maturity ranking,' or asks questions like 'how do these companies compare on AI,' 'who is ahead on AI,' 'rank these firms by AI capability,' or 'score Company X on AI.' This skill handles everything from a quick 3-company comparison to a deep 15-company competitive intelligence analysis across any industry."

skill-discovery

by weskao

Daily discovery of new Claude Code skills AND adjacent AI/agent tools. Detects the host project home (e.g. ~/.claude or ~/.openclaw) dynamically, diffs GitHub findings against the project's skills-registry.yaml, scores candidates, writes a shortlist to skill-candidates.yaml, and sends a Telegram message for user approval. Triggered on a daily cron via /schedule, or manually invoked. Also handles user replies that approve installation of candidates.

Ojas Engineering Skill

by Codered-Vigneshvar

Prefer the simplest thing that satisfies the invariants. Ask before introducing a new dependency, a new service, or a new database.