aaaaqwq/claude-code-skills
152 skills10.5k installs12.5k starsGitHub
Install
npx skills add https://github.com/aaaaqwq/claude-code-skillsSkills in this repo
1Multi Search Enginemulti-search-engine is an agent skill for aggregating search queries across seventeen configured engines in one workflow. It supports simultaneous domestic China engines including Baidu, Bing CN, 360, Sogou, Toutiao search, WeChat Sogou, and Jisilu alongside global engines such as Google, Google HK, DuckDuckGo, Yahoo, Startpage, Brave, Ecosia, Qwant, and WolframAlpha. Use cases include comparing results across engines, running regional CN versus Global searches, and retrieving multi-source web answers from a single config-driven setup. Configuration lives in config.json with URL templates using a keyword placeholder for query substitution. The skill documents supported engine lists by region and points to config.json as the authoritative engine configuration list. Developers reach for it when they need parallel multi-engine retrieval rather than a single default search provider, especially for cross-region research where domestic and international indexes diverge. Related files include config.json for the search engine list and URL template definitions that agents load before issuing aggregated queries.2.7kinstalls2Web Scraping Automationweb-scraping-automation is a coding-agent skill for automating website data extraction and API integration, covering site structure analysis, REST and GraphQL calls, crawler script generation, parsing, anti-bot handling, and scheduled data collection. Allowed tools include Bash, Read, Write, Edit, WebFetch, and WebSearch, enabling end-to-end script creation inside the agent session. The skill documents Playwright-based browser automation with mandatory Chrome and Selenium process cleanup after browser tasks, plus patterns for headless Chromium scraping. Developers reach for web-scraping-automation when they need product catalogs scraped, news feeds collected on a schedule, API responses parsed into structured records, or anti-scraping workarounds prototyped quickly. It suits backend and data-integration tasks where repeatable scripts—not manual copy-paste—must land in the repository.590installs3Wechat Channelwechat-channel is an integration skill in aaaaqwq/claude-code-skills that connects WeChat users to an OpenClaw AI agent gateway through a Wechaty bridge using the PadLocal protocol. The architecture routes private chats, group chats, @mention detection, and image or file transfers between WeChat clients and OpenClaw so users can chat with an AI assistant directly inside WeChat. Developers reach for wechat-channel when WeChat messages must trigger AI responses or when OpenClaw needs to send replies back into WeChat conversations. The skill requires Bash, Read, Write, and Edit tool access to configure and run the bridge components.509installs4Project Managementproject-management is a bilingual (Chinese/English) Claude Code skill from aaaaaqwq/claude-code-skills focused on product requirements and delivery planning. It writes PRDs, builds project schedules, splits features into executable tasks, defines user stories with acceptance criteria, and flags technical, market, adoption, and schedule risks. Priority tiers use P0 Must Have, P1 Should Have, and P2 Nice to Have for MVP scoping. Allowed tools include Read, Write, Edit, TodoWrite, and WebSearch for competitor or market context. Developers and tech leads invoke it when translating vague feature ideas into structured specs agents can implement, planning MVPs, or running agile breakdown sessions before sprint coding starts.467installs5Wecom Automationwecom-automation is a skill from aaaaqwq/claude-code-skills that helps developers wire WeCom (WeChat Work) into product integrations. It covers automated messaging, team notifications, and group workflow triggers so engineering teams can push alerts, onboarding messages, or ops updates through enterprise WeChat instead of building ad-hoc scripts. Reach for wecom-automation when a SaaS or internal tool must notify Chinese enterprise users on WeCom during deployment hooks, billing events, or support escalations. The skill assumes familiarity with WeCom bot or app credentials and enterprise approval flows typical in China-based org stacks.425installs6Feishu AutomationFeishu-automation skill teaches Claude Code to integrate Lark/Feishu via bots, webhooks, and APIs for sending messages, syncing documents, routing approvals, and automating team workflows inside Chinese enterprise collaboration environments.367installs7Memory Hygienememory-hygiene is an agent skill from aaaaqwq/AGI-Super-Team (forked from openclaw memory-hygiene) for keeping Clawdbot vector memory lean. It audits stored entries with memory_recall query="*" limit=50, wipes bloated LanceDB data via rm -rf ~/.clawdbot/memory/lancedb/, and reseeds intentional facts from MEMORY.md using memory_store with category preference, fact, or decision and importance 0.7–1.0. The skill disables autoCapture in the memory-lancedb plugin config—the main source of junk memories—while keeping autoRecall enabled, and documents a monthly cron schedule (0 4 1 * *) for automated maintenance. Developers reach for memory-hygiene when agent token usage spikes from irrelevant auto-recalls or vector memory fills with heartbeat noise and transient chat logs.365installs8Seo Content WritingTeaches Claude Code to produce SEO-oriented written content—articles, landing sections, and product narratives—with deliberate keyword placement, semantic headings, meta descriptions, and scannable structure so new pages can compete in organic search without sacrificing clarity.225installs9Task Statustask-status gives Claude Code a lightweight project-management layer for reporting what is in progress, completed, or blocked during implementation. It helps solo builders and agent loops maintain alignment with planned tasks without switching to external trackers.163installs10Uml Diagram DesignHelps Claude Code author clear UML diagrams—class, sequence, component, and state charts—that document domain models, service interactions, and lifecycle rules, giving developers and agents a shared visual contract before and during implementation.122installs11Media Auto Publishermedia-auto-publisher automates preparing and pushing media—posts, clips, and articles—to multiple distribution endpoints with consistent metadata, helping growth teams maintain cadence and reduce handoffs between creation tools and social or CMS platforms.113installs12PptxClaude Code pptx skill for creating and editing PowerPoint presentations with consistent layouts, speaker notes, visuals, and branding for pitches, trainings, and stakeholder slide decks.112installs13Content FactoryContent-factory skill from claude-code-skills guides Claude Code through scalable content production: topic batches, templated drafts, variations for multiple channels, and editorial workflows that sustain marketing and SEO growth after product launch.107installs14Docx PerfectClaude Code skill for producing publication-quality Word documents—applying consistent styles, tables, headings, and layout fixes for reports, memos, proposals, and other business content deliverables.104installs15Proactive Agent LiteThis skill adds proactive behavior patterns to an AI agent, including a memory architecture that flushes context before the window fills, reverse prompting that surfaces unrequested ideas, and self-healing patterns that recover from errors. A developer uses it to make an agent anticipate needs and improve across interactions instead of just following tasks. It is a lightweight version aimed at OpenClaw agents.99installs16Chrome Automationchrome-automation is a Claude Code skill for automating the Chrome browser with tools like Puppeteer, Playwright, Selenium, and Cypress. A developer uses it for web automation, E2E testing, screenshots and PDF generation, form filling, and dynamic-page data scraping. It provides TypeScript examples for navigation, element operations, scraping, performance tracing, request interception, and mobile emulation. Documentation is primarily in Chinese.93installs17Skill Finder Cnskill-finder-cn is a skill that helps users discover and install skills from the ClawHub registry. When a user asks what skill can do a given task, it extracts keywords, runs clawhub search, and recommends and installs matching skills. It is a bilingual Chinese/English wrapper around the clawhub CLI. It requires the clawhub binary to be available.91installs18Feishu Channelfeishu-channel is a Claude Code skill that connects Feishu (Lark) to OpenClaw as a two-way message channel. It receives user messages through webhook event subscriptions and sends replies, supporting direct messages, group chats, @-mention detection, card messages, and file transfer. A developer uses it to build an AI chatbot that responds to Feishu messages. It is distinguished from feishu-automation, which handles platform automation like bitables and docs rather than the message channel.89installs19Email Manageremail-manager is a Claude Code skill that unifies multiple IMAP mailboxes such as Gmail and QQ Mail. It checks mail on a schedule, generates AI summaries of unread messages, and drafts replies that require user confirmation before sending. A developer uses it to triage inbox volume without manually opening each account. It stores credentials with pass and gates sending and deletion behind explicit permission tiers.88installs20Agent Team Orchestrationagent-team-orchestration is a Claude Code skill that provides a playbook for running multi-agent teams with defined roles, a task lifecycle, handoff protocols, and review gates. A developer uses it to set up teams of two or more agents, route tasks through states, and enforce quality reviews between builders and reviewers. It includes reference files for team setup, task lifecycle, communication, and workflow patterns.87installs21Wecom Cs Automationwecom-cs-automation is a Claude Code skill for building an AI customer-service assistant on WeChat Work (WeCom). It auto-accepts friend requests, answers questions using a PostgreSQL pgvector knowledge base with an LLM, and escalates low-confidence questions to a human via Telegram. A developer uses it to stand up an automated support bot with a FastAPI callback server. It ships a schema, import scripts, and workflow handlers.86installs22Facebook Automationfacebook-automation is a Claude Code skill that drives Facebook operations through Composio's Facebook toolkit via Rube MCP. It manages pages, creates and schedules posts, replies to comments, and pulls page and post insights. A developer or marketer uses it to run Facebook page tasks from the agent without hand-coding the API. It requires an active Facebook OAuth connection and reads current tool schemas before each workflow.85installs23Github Automationgithub-automation is a skill for automating GitHub operations through the GitHub MCP. It covers repository and file management, pull requests, issues, branches, code search and CI/CD workflows. It instructs the agent to prefer mcporter call github.push_files over git push so pushes work without a personal access token or SSH config. A developer uses it to push code, open PRs and manage issues programmatically.76installs24Project PlannerThis skill helps plan a project's execution path, decomposing work into a work-breakdown structure with dependencies, milestones, timelines, and resource assignments. A developer or project lead uses it when starting, scheduling, or replanning a project to keep delivery on time and high quality. It also produces a risk register, communication plan, and quality/acceptance checklists, distinct from product requirements analysis.76installs25Backend Developmentbackend-development is a Claude Code skill (documented mostly in Chinese) that acts as a backend generalist, routing to language-specific sub-skills for Python, Node.js, Go, and Java. A developer uses it to choose a framework, database, and auth scheme, then follow general backend best practices like REST conventions, rate limiting, and graceful shutdown. It bundles per-language SKILL.md files for deeper guidance.73installs26Rss Automationrss-automation is a skill for aggregating and monitoring RSS and Atom feeds. It parses feeds with Python feedparser, filters entries, tracks seen versus new items with hashed IDs, and pushes new items to channels like Telegram or Slack. A developer uses it to set up automated content monitoring on a cron schedule. It includes rate-limiting and persistent-state guidance so checks survive restarts.73installs27Playwright AutomationThis skill uses Playwright to automate a headless browser for data scraping, form filling, UI testing, screenshots, and PDF generation. A developer uses it for unattended cron jobs where no human is present to open a browser. It documents sync and async APIs, element selectors, wait strategies, and anti-scraping tactics, and requires killing Chrome/Chromium processes after each run to avoid resource leaks.61installs28Web Searchweb-search is a Claude Code skill that gives an agent internet search and web page fetching. It routes each request to WebFetch for single pages, Tavily for AI-optimized search, or Firecrawl for structured scraping and batch crawls. A developer uses it when the agent needs real-time data or page content. The skill documents when to pick each tool and the exact shell commands to call.61installs29Afrexai Personal Financeafrexai-personal-finance is a Claude Code skill that runs a personal finance system covering budgeting, debt elimination, investing, tax optimization, and net worth tracking. A user invokes it to score their financial health, build a 50/30/20 budget, or pick a debt-payoff strategy. It ships checklists, YAML templates, and benchmark tables with no external dependencies.60installs30Frontend DevelopmentDespite the listing name, this skill's SKILL.md is a paid-advertising audit and optimization tool named 'ads'. It performs multi-platform ad-account audits across Google, Meta, LinkedIn, TikTok and Microsoft Ads, checking conversion tracking, creative, budget allocation, bidding and compliance. It detects the business type and outputs an Ads Health Score (0-100) with a graded, prioritized action plan. A developer or marketer uses it to review and improve PPC campaign performance.60installs31Skill Creatorskill-creator is a meta skill that guides creating new skills or updating existing ones for Claude Code. It documents the SKILL.md folder structure, frontmatter format, and authoring best practices, and ships scripts to initialize, package, and validate a skill. A developer uses it when authoring a skill that extends the agent with new workflows or tool integrations. The SKILL.md body is written in Chinese.60installs32Notion Automationnotion-automation is a Claude Code skill that automates Notion operations such as creating and editing pages, configuring and querying databases, updating blocks, and managing project tracking boards. It defines standard workflows for page creation, database setup, content updates and Kanban task tracking. A developer uses it to script repeatable Notion management through the notionApi MCP tools.59installs33Daily RhythmThis Claude Code skill sets up a daily and weekly planning routine that automates morning briefings, evening wind-down prompts, sleep nudges, and weekly reviews. It syncs data from Google Tasks, optional Stripe ARR, and a calendar ICS feed, then compiles a formatted morning brief. A solo developer uses it to keep a structured daily rhythm and track priorities via cron jobs.56installs34Healthcare Monitorhealthcare-monitor is a skill that monitors Chinese healthcare companies for undisclosed funding signals. It periodically scrapes business-registry data (Tianyancha/Qichacha) via Playwright, does incremental change detection, and scores signals like new institutional shareholders or registered-capital increases into confidence-weighted funding alerts. It pushes alerts to Telegram and Feishu and generates daily and weekly reports. It is used by media, VC/PE and pharma BD teams to spot funding leads early.55installs35Crypto Portfolio Managementcrypto-portfolio-management is a Claude Code skill that guides an agent in helping users build and manage crypto portfolios. It covers asset allocation by risk profile, rebalancing strategies, risk management, performance metrics, and tax-loss harvesting. A developer uses it when advising on portfolio construction, rebalancing, or DeFi yield.54installs36De Ai IfyThis Claude Code skill rewrites AI-generated text into a more natural, human-sounding style. The guide (written in Chinese) gives principles like breaking rigid structure, adding colloquialisms and personal voice, and mixing sentence lengths, with before/after examples and a self-check list. A writer uses it to remove the mechanical tone from AI drafts.49installs37Docker DeploymentThis Claude Code skill covers deploying Docker web applications with an Nginx HTTPS reverse proxy and Cloudflare Tunnel for secure external access. It documents SSL certificate mounting, named-volume data persistence, and common HTTPS and data-loss troubleshooting. A developer uses it when containerizing and exposing a web app over HTTPS.47installs38Model Fallbackmodel-fallback is a Claude Code skill that provides automatic model downgrade and failover. When a primary model request fails, times out, or hits a rate limit or quota, it switches to a backup model to keep service running. It supports multi-provider, priority-based model selection with health monitoring, auto-retry, and error recovery, and switches back once the primary recovers. A developer uses it to keep an AI agent reliable across model outages. The SKILL.md is written in Chinese.47installs39Openclaw Inter Instanceopenclaw-inter-instance is a Claude Code skill for communicating between OpenClaw instances and agents. It ranks communication methods by reliability, from real-time sessions_send agent-to-agent messages to nodes.run remote execution and file-level fallback, and explains which methods are blocked. A developer uses it to coordinate multi-agent tasks and cross-machine execution across OpenClaw instances.47installs40N8n Workflow AutomationThe files bundled under n8n-workflow-automation are actually a paid-advertising audit skill named ads. It audits ad accounts across Google, Meta, LinkedIn, TikTok and Microsoft, detects the business type, scores account health from 0 to 100, and outputs a prioritized optimization plan. A marketer or growth developer uses it to review PPC accounts and find optimization wins.44installs41Figma Ui DesignDespite the listing name, this skill's SKILL.md is an 'ads' skill for multi-platform paid advertising audit and optimization. It audits Google, Meta, LinkedIn, TikTok, and Microsoft ad accounts, checks conversion tracking and creative quality, and reviews budget and bidding. A marketer uses it to score ad accounts and get a prioritized action plan. It orchestrates 12 sub-skills and 6 subagents and detects the business type before auditing.42installs42Doc CoauthoringThis Claude Code skill guides a structured workflow for co-authoring documentation such as technical specs, proposals, decision docs, and guides. The guide (written in Chinese) breaks the work into context transfer, content iteration, and reader verification phases. A developer uses it to efficiently draft and refine structured documents with the agent.41installs43Feishu Doc Optimizerfeishu-doc-optimizer is a Claude Code skill that improves the structure, formatting, and readability of Feishu cloud documents. It reads a document by token via the Feishu API, restructures the content, and replaces it using Playwright browser automation because the API cannot delete document blocks directly. A developer or writer uses it to clean up and standardize Feishu doc layout. It applies a defined convention of section numbering, emoji markers, separators, and aligned tables.41installs44Senior Architectsenior-architect is a skill for system-design and architecture analysis. It ships Python tools to generate architecture diagrams from a codebase, analyze dependencies for coupling and circular references, and assess a project's architectural patterns and code smells. A developer uses it to design system architecture, write ADRs, choose a database, and review system design. It supports npm, Python, Go, and Rust package manifests.41installs45Electron App DevThis Claude Code skill helps build cross-platform Electron desktop apps using electron-vite, TypeScript, and React. The guide (written in Chinese) enforces secure IPC via contextBridge with channel whitelists and documents window management, native features, packaging, and common pitfalls like white screens and memory leaks. A developer uses it to scaffold and harden an Electron app.39installs46Mcp Installermcp-installer is a Claude Code skill that finds and installs MCP (Model Context Protocol) servers. It searches GitHub for MCP projects, extracts the correct npx install command, and adds the server configuration to the user's ~/.claude.json without breaking existing entries. When a server needs an API key, it inserts a placeholder and explains how to get the key. A developer uses it to add a new MCP tool to Claude Code without editing config by hand. The SKILL.md is written in Chinese.39installs47Arxiv Automationarxiv-automation is a Claude Code skill for searching, monitoring, and analyzing academic papers on arXiv. A developer or researcher uses it to query papers by keyword, author, or category, track new submissions via RSS, and download PDFs to summarize abstracts. It includes a ready Python search function and notes arXiv's rate limits.38installs48Openclaw Memory Enhanceropenclaw-memory-enhancer is a Claude Code skill that gives OpenClaw agents long-term memory with semantic vector search. It auto-loads files from a memory directory, recalls relevant context during conversations, and stores everything locally. It ships an edge build under 10MB with zero dependencies for Jetson and Raspberry Pi, plus a higher-accuracy standard build using sentence-transformers.38installs49Csv Pipelinecsv-pipeline is a Claude Code skill that processes, transforms, analyzes, and reports on tabular data files. A developer uses it to filter rows, join datasets, compute aggregates, convert formats, or deduplicate CSV, TSV, and JSON data. It supplies shell one-liners and dependency-free Python functions for ETL workflows.37installs50Multi Agent Architecturemulti-agent-architecture is a Claude Code skill for designing multi-agent systems in OpenClaw. It lays out a main agent that dispatches to specialized agents (news, code, research, quick, batch, monitor), each with its own model, system prompt, tools, and channel bindings. It shows how to register agents and implements a smart spawn system that routes tasks by type, complexity, time-sensitivity, and cost. A developer uses it to build a coordinated multi-agent setup. The SKILL.md is written in Chinese.37installs51Multimodal Genmultimodal-gen is a Claude Code skill that turns text prompts into images or videos. It detects image or video generation intent and calls a multimodal-agent through sessions_spawn, choosing a model from aliases like flux, imagen, dalle, sora and kling. A developer uses it when they want text-to-image or text-to-video output without wiring each provider by hand.36installs52Security Monitorsecurity-monitor is a skill that runs continuous security monitoring on a Clawdbot deployment. It watches for intrusions, unusual API calls, credential-usage patterns, port scans, process anomalies, and file changes, then alerts via console, JSON logs, or Telegram. A developer runs it as a daemon (via PM2 or systemd) to keep watch after deployment. It requires no external dependencies and pairs with a separate one-time security-audit skill.36installs53Auth Managerauth-manager is a Claude Code skill (documented in Chinese) that manages web login and session state across platforms using OpenClaw's built-in browser and a single shared Chrome profile. A developer uses it to periodically check whether logins are still valid, authorize new platforms, and let other agents inherit those sessions. It prefers fast curl/CLI checks and falls back to browser snapshots, with a cron job that alerts on expired sessions.35installs54Context Managercontext-manager is a Claude Code skill that compresses long OpenClaw agent sessions. A developer uses it when context usage approaches 70-80% to summarize, back up, reset, and re-inject a compressed session. It ships a compress.sh script with list, status, summarize, and threshold commands.35installs55DocxThis Claude Code skill works with Word .docx files to create, edit, and analyze documents, including tracked changes and comments. It uses pandoc for text extraction, docx-js to create new documents, and a Python OOXML library to edit existing ones, with a redlining workflow for reviewing others' documents. A developer uses it to programmatically manipulate professional Word documents.35installs56Permission ManagerThis skill configures Claude Code global or project tool-execution permissions so MCP and other tool commands are auto-approved instead of prompting each time. A developer uses it to add commands to allowedTools in ~/.claude.json using a Node script that merges and deduplicates entries. It includes preset tool lists for Playwright, Figma, and GitHub MCP servers and reminds the user to restart Claude Code.35installs57Mcp Managermcp-manager is a Claude Code skill that manages MCP (Model Context Protocol) servers. It runs periodic health checks on connection status, response time, and error rate, and alerts when a server is unavailable. It auto-disables idle servers and re-enables them when needed, and answers natural-language questions about each MCP's capabilities and limits. A developer uses it to keep a set of MCP servers healthy and resource-efficient. The SKILL.md is written in Chinese.34installs58Team Coordinatorteam-coordinator makes the agent act as an executive that breaks a user task down and assigns it to the most suitable specialist worker agents, coordinating parallel work and reviewing the output. A developer uses it to route complex, multi-domain requests across a team of agents instead of doing the work directly. It defines a capability matrix and a task-routing decision flow.34installs59Env Setupenv-setup is a Claude Code skill that syncs Claude Code and OpenClaw configuration from a GitHub repository to the local machine. It copies output-styles, the global CLAUDE.md, MCP server config, agent configs, and plugins into their target directories. A developer uses it to keep several machines consistent, restore a setup on a new computer, or share config across a team. MCP config is merged rather than overwritten, and it warns against committing real API keys.33installs60File Cleanerfile-cleaner is a Claude Code skill that scans directories for large files and junk files such as temp files, caches, logs, and build artifacts. It offers an interactive cleanup interface, a dry-run mode, and a generated cleanup script. A developer uses it to reclaim disk space and clear build and cache clutter from projects. It auto-excludes system, version-control, and Python environment directories to avoid deleting important files.33installs61Linux Service Triagelinux-service-triage is a Claude Code skill that diagnoses common Linux service failures. It works from logs and status output to classify the failure (config error, permission denied, port conflict, upstream unreachable, DNS mismatch) across systemd/PM2, file permissions, Nginx reverse proxy, and DNS. A developer uses it when a server app is failing, unreachable, or misconfigured. It outputs a triage report and, only with approval, the exact fix commands.33installs62Browser Usebrowser-use is a Claude Code skill for AI-driven browser automation built on the browser-use Python library, where an LLM reads pages and decides the next action instead of following a fixed script. A developer uses it for complex or dynamic browser tasks that need intelligent decisions, such as extracting structured data from a site. It documents LLM configuration, Chrome sessions, login-state persistence, and token-reduction settings. Documentation is primarily in Chinese.32installs63Firecrawlfirecrawl is a Claude Code skill that wraps the Firecrawl API for web scraping and data extraction. It scrapes single pages including dynamic content, extracts structured data against a schema, crawls sites in batch, and searches then scrapes. A developer uses it to pull web content and structured data into a workflow via a bundled shell script. It reads the Firecrawl API key from the pass store and notes plan-based usage limits.32installs64Token Guardtoken-guard monitors an AI agent's token usage and costs, sets daily budgets, and auto-downgrades the model when a hard limit is hit. A developer running an agent around the clock uses it to cap spend and avoid burning Opus-level tokens overnight. It ships a shell script for status, budget-setting, cost estimation, and usage history.31installs65Api Provider Setupapi-provider-setup is a Claude Code skill (documented in Chinese) for adding and configuring third-party API relay providers into OpenClaw. A developer uses it to register custom model endpoints in openclaw.json, set aliases and default/fallback models, and sync cached auth profiles after changing keys. It supports Anthropic-compatible and OpenAI-compatible provider formats.30installs66Frontend Designfrontend-design is a Claude Code skill that guides building distinctive, production-grade frontend interfaces. It has the agent commit to a bold aesthetic direction and then implement working code across HTML/CSS/JS, React or Vue. It steers typography, color, motion, layout and backgrounds so generated UI avoids generic AI-slop aesthetics. A developer uses it when building web components, pages, landing pages or dashboards.30installs67Xiaohongshu Workflowxiaohongshu-workflow is a Claude Code skill for operating a Xiaohongshu (RED) account end to end. It covers MCP server setup and cookie login, publishing image and video notes, tracking trending topics, managing comment replies, and generating analytics reports. A developer uses it to automate content marketing on Xiaohongshu. It bundles many shell scripts and reference docs built on the xiaohongshu-mcp project.30installs68Bat Catbat-cat is a Claude Code skill that documents how to use the `bat` command-line tool, a cat replacement with syntax highlighting, line numbers, and Git integration. A developer uses it to preview files, view logs with highlighting, show specific line ranges, and pipe highlighted output to other shell tools. It also covers themes, config defaults, and integration with ripgrep, fzf, and man pages.29installs69Performing Security Code ReviewThis skill invokes a security-agent plugin to review code for vulnerabilities like SQL injection, XSS, authentication flaws, and insecure dependencies. A developer uses it when auditing a codebase or a specific query before shipping. It returns a structured report listing each finding with severity, the affected code, and recommended fixes.29installs70Skill Searchskill-search (skill-search-optimizer) is a meta skill for making agent skills discoverable on the ClawdHub/MoltHub registry. It explains that the registry uses vector-based semantic search, which fields are indexed, and how to write descriptions and keywords that rank well. A skill author uses it when a published skill is not being found or when writing a description for maximum discoverability. It includes description formulas and length guidance.29installs71Data AnalystThis Claude Code skill turns the agent into a data analyst that writes and runs SQL, processes CSV and Excel data with pandas, cleans data, and generates reports and visualizations. It ships reusable query templates for cohort, funnel, and time-based analysis plus a data-quality audit checklist. A developer uses it to explore a database or spreadsheet and produce statistics and insights.28installs72Twitter Automationtwitter-automation automates Twitter/X operations (posts, search, user lookups, bookmarks, lists, and media) through Composio's Twitter toolkit via Rube MCP. A developer uses it to create or delete tweets, search recent posts, and manage bookmarks or lists from an agent. It documents each workflow's tool sequence, key parameters, and pitfalls like the 280-character limit and non-idempotent posting.28installs73Openclaw Config Helperopenclaw-config-helper is a Claude Code skill that enforces a safe process for editing OpenClaw agent-gateway configuration. It requires checking the config schema and official docs, showing the planned change for confirmation, then applying it with config.patch and verifying. A developer uses it when changing OpenClaw channels, bindings or agent settings without breaking the system.27installs74Tavilytavily wraps the Tavily API to give an agent AI-optimized web search and content extraction. A developer uses it when the agent needs to search the internet, fetch real-time data, or find the latest information. It runs a bundled tavily.sh script for basic search, deep search, and search-and-extract, reading its API key from pass.27installs75Analyzing Options Flowanalyzing-options-flow is a Claude Code skill for tracking crypto options flow to gauge institutional positioning and market sentiment. A developer invokes it with phrases like 'track options flow' to query crypto market and blockchain data via APIs such as CoinGecko, Etherscan, Dune, and The Graph. It aggregates prices, volumes, and liquidity metrics across exchanges.26installs76Deep ResearchThis Claude Code skill produces cited research reports by searching the web with firecrawl and exa MCP tools and synthesizing findings with source attribution. It follows a six-step workflow from clarifying the goal to writing a structured report, and can parallelize across subagents. A developer uses it for competitive analysis, technology evaluation, or any question needing multi-source evidence.26installs77Vibe Code Auditorvibe-code-auditor audits rapidly generated or AI-produced code for structural flaws, fragility, and production risks. A developer uses it to check whether prototype-quality code that works is actually robust and maintainable before productionizing it. It evaluates seven dimensions and outputs a severity-ranked report of real findings with locations and minimal recommended fixes.26installs78Xiaomo Assistant Templatexiaomo-assistant-template is a Claude Code skill that scaffolds a new OpenClaw AI assistant from a set of template files. It provides SOUL.md, USER.md, IDENTITY.md, HEARTBEAT.md, and MEMORY.md to define the assistant's personality, user profile, identity, scheduled checks, and long-term memory. A developer uses it to stand up a personalized assistant quickly. It is adapted from the xiaomo-starter-kit.25installs79Defi Yield StrategiesDespite the listing name, the bundled SKILL.md and reference files describe a paid-advertising audit and optimization skill. It analyzes Google, Meta, LinkedIn, TikTok, and Microsoft ad accounts, orchestrates sub-skills and subagents, detects the business type, and outputs an Ads Health Score with a prioritized action plan. A marketer uses it to audit and optimize campaigns.24installs80Prompt OptimizerThis skill evaluates prompt quality and rewrites prompts using 58 prompting techniques such as chain-of-thought, few-shot, and role-play. A developer uses it to improve a prompt's clarity, specificity, and structure or to generate optimized variations. It scores prompts across defined quality dimensions and can run bundled Python scripts for repeatable evaluation and optimization.24installs81Defi Risk AssessmentThis Claude Code skill gives the agent a framework to assess DeFi protocol risk before a user deposits funds. It scores five risk categories (smart contract, economic, centralization, liquidity, regulatory) with a weighted formula and includes a red-flags checklist and due-diligence steps. A user invokes it to compare protocols and spot warning signs.23installs82Agent Patternsagent-patterns is a Claude Code skill that defines structured message formats and communication protocols for coordinating multiple agents. A developer uses it to format SPAWN REQUEST messages, produce standardized AGENT REPORT status reports, and run a spawn-monitor-merge coordination workflow. It targets multi-agent sprint systems where agents must communicate traceably.22installs83Content Source Aggregatorcontent-source-aggregator is a Claude Code skill that collects real-time trends and hot topics from 16+ platforms. A team uses it as the data-collection stage of a content pipeline, feeding trending items to research and content agents. It normalizes results into a dated JSON hot-topic pool with engagement metrics.22installs84Generating Trading Signalsgenerating-trading-signals is a skill that generates composite trading signals from technical indicators. It runs Python scripts that analyze price action with 7 indicators (RSI, MACD, Bollinger Bands, trend, volume, stochastic, ADX) and produces BUY/SELL signals with confidence scores and risk-management levels. It can scan watchlists, filter and rank opportunities, and export results to JSON. A developer uses it to screen crypto or stock assets for trade entries.22installs85Research Engineerresearch-engineer is a persona skill that makes Claude Code behave like an academic research engineer focused on correctness over politeness. It critiques a flawed premise first, refuses to invent libraries or fabricate complexity bounds, and writes complete, compilable, tested implementations. A developer uses it when working on performance-critical or algorithmically hard code where rigor and no placeholders matter. It includes a language-selection matrix and an optimization tier list.22installs86Risk Metrics Calculationrisk-metrics-calculation is a skill for measuring portfolio risk in code. It covers Value at Risk, Conditional VaR (Expected Shortfall), Sharpe and Sortino ratios, and drawdown analysis, with Python patterns built on numpy, pandas, and scipy. A developer uses it when building risk dashboards, setting position sizes, implementing risk limits, or doing regulatory reporting. A detailed implementation playbook holds the code samples.22installs87Skill Search Optimizerskill-search-optimizer teaches how to write skill descriptions and content so they rank in the ClawdHub/MoltHub registry's semantic search. A developer uses it when a published skill is not appearing in relevant searches or when writing a new description for discoverability. It matters because the description field is the primary indexed field that determines whether other users find and install the skill.21installs88Backtesting Frameworksbacktesting-frameworks is a Claude Code skill for building trading-strategy backtesting systems that avoid common biases and produce reliable performance estimates. A developer uses it to define a hypothesis and universe, build point-in-time data pipelines and realistic cost models, and run walk-forward testing. An implementation playbook covers backtesting biases and patterns in detail.20installs89Marketing Ideasmarketing-ideas is a Claude Code skill that helps developers pick and prioritize marketing tactics for SaaS and software products. It draws on a library of 140 marketing ideas and scores candidates with a Marketing Feasibility Score across impact, effort, cost, speed-to-signal, and fit. Instead of dumping long lists, it recommends the top 3-5 ideas with first steps, success metrics, and risks. A developer uses it to decide what to try now versus defer.20installs90Remembering ConversationsThis skill searches previous Claude Code conversations for facts, patterns, decisions, and context using semantic similarity or exact text matching. A developer uses it when a past discussion is referenced, when debugging a familiar issue, or when looking for prior architectural decisions. It runs a search-conversations tool with vector, text, or both modes plus date and limit filters, and returns matched exchanges with similarity scores.20installs91Billing Automationbilling-automation is a Claude Code skill for building automated billing systems, covering recurring billing, invoice generation, dunning for failed payments, proration, and tax calculation. A developer uses it when implementing SaaS subscription billing or automating invoicing and renewal flows. It walks through subscription lifecycle states, billing cycle processing, and points to a detailed implementation playbook with Python examples.19installs92Content Marketercontent-marketer is a Claude Code skill that acts as a content marketing strategist across channels. A developer or marketer uses it for content creation, SEO, social media, email automation, distribution, and performance analytics. It bundles capabilities and an implementation playbook covering the full content marketing lifecycle.19installs93Kubernetes Deploymentkubernetes-deployment is a Claude Code workflow for deploying applications to Kubernetes. It walks through seven phases: container prep, K8s manifests, Helm charts, service mesh, security, observability, and GitOps deployment. Each phase names sub-skills to invoke and gives copy-paste prompts, ending with a quality-gate checklist. A developer uses it to go from a Dockerfile to a monitored, GitOps-deployed cluster.19installs94Openrouter Usageopenrouter-usage is a Claude Code skill that tracks OpenRouter API spending for an OpenClaw agent. It reports credit balance, per-model cost and token breakdowns, and spending projections by parsing local session logs. A developer uses it to check how much they are spending on OpenRouter by day, week or month without leaving the agent.19installs95React Component GeneratorThis skill provides automated assistance for generating React components within frontend development. A developer triggers it when they need help creating React components or applying frontend patterns. It generates component code and configurations and covers React, Vue, CSS, accessibility, and performance topics, though its documentation is thin and mostly generic.19installs96Sergei Mikhailov Tg Channel Readersergei-mikhailov-tg-channel-reader is a skill that reads posts and comments from Telegram channels using MTProto (Pyrogram or Telethon). It fetches recent messages by time window from public or subscribed private channels, can pull discussion replies, and supports an unread-only mode for daily digests. A developer uses it to monitor or summarize Telegram channels. It requires TG_API_ID and TG_API_HASH credentials and a session file from my.telegram.org.19installs97Subagent Driven Developmentsubagent-driven-development executes an implementation plan by dispatching a fresh subagent for each independent task and running a two-stage review after each: spec compliance then code quality. A developer uses it to run a plan in one session without polluting the coordinator's context. It keeps each subagent focused on isolated context and iterates faster than a human-in-the-loop.19installs98Agent BrowserAgent Browser is a Claude Code skill that wraps the agent-browser headless-browser CLI so an AI agent can drive a real browser through structured commands. A developer uses it to navigate pages, snapshot interactive elements, click and fill by ref, and capture screenshots or PDFs. It also covers form filling, waits, cookies/storage, and browser settings for automated web interaction.18installs99Geo Agentgeo-agent is an automated GEO (Generative Engine Optimization) skill that boosts a brand's visibility in AI search engines. It manages keywords, researches real competitors, and generates comparison, ranking and Q&A articles that place the target brand prominently. It auto-publishes to Chinese content platforms via Playwright and monitors indexing in AI engines like Doubao, Qwen, DeepSeek and Perplexity. A developer or marketer uses it to run an end-to-end AI-search content campaign.18installs100Podcast GenerationThis skill generates podcast-style audio narratives from text using Azure OpenAI's GPT Realtime Mini model over a WebSocket. A developer uses it to add text-to-speech or audio narrative features to an app. It streams PCM audio chunks and a transcript, converts PCM to WAV, and returns base64-encoded audio for frontend playback, with six selectable voices.18installs101Duckdb EnThis Claude Code skill is a DuckDB CLI reference for SQL analysis, data processing, and file conversion. It shows how to query CSV, Parquet, and JSON files directly, convert between formats, use output flags and dot commands, and pipe data through stdin/stdout. A developer uses it for quick command-line data analysis and format conversion.17installs102PolymarketThis skill queries Polymarket prediction markets for live odds, prices, order books, events, and search results, treating prices as implied probabilities. A developer uses it to answer market-sentiment or win-probability questions and, with the Polymarket CLI and a wallet, to read portfolios, leaderboards, and place trades. Core read commands work without auth; advanced portfolio, analytics, and trading commands require the CLI binary and a private key.17installs103Startup Metrics Frameworkstartup-metrics-framework is a reference for tracking, calculating, and optimizing key startup and SaaS performance metrics from seed through Series A. A founder or operator uses it when working out CAC, LTV, unit economics, burn multiple, or Rule of 40. It provides formulas and benchmark targets in a bundled implementation playbook.17installs104Ab Test Setupab-test-setup helps plan, design and analyze A/B and multivariate experiments that produce statistically valid results. It provides a hypothesis framework, sample-size reference tables, metric selection (primary, secondary, guardrail), traffic allocation, and a significance analysis checklist. A developer or marketer uses it to design a conversion experiment and avoid pitfalls like stopping early.16installs105Company Analyzercompany-analyzer is a Claude Code skill for investment research on public companies using 8 specialized analysis frameworks. A developer or investor runs it on a ticker to classify company phase, score financial metrics, assess AI and strategic moats, gauge sentiment, and analyze growth, business model, and risk. It runs frameworks in parallel with caching and cost tracking, pulling data from SEC EDGAR and Alpha Vantage, then synthesizes an investment view.16installs106Competitor Alternativescompetitor-alternatives is a Claude Code skill that generates competitor comparison and alternative landing pages for SEO and sales enablement. A developer uses it when writing 'X vs Y', '[Competitor] alternative', or comparison pages that need to rank for competitive search terms. It supplies four page formats, section templates, and a centralized competitor-data architecture.16installs107Ct Monitorct-monitor is a Claude Code skill that acts as a crypto intelligence analyst over the ctmon.xyz API. A user asks it for market sentiment, token research, KOL activity, alpha signals, or AI briefings, and it queries tweets, news, and prices. It requires a CT_MONITOR_API_KEY and enforces strict no-fabrication rules.16installs108Geo Content Optimizergeo-content-optimizer is a skill that optimizes written content for Generative Engine Optimization so AI systems are more likely to cite it. It applies a prioritized CORE-EEAT checklist covering direct answers, FAQ schema, data tables, JSON-LD and authority signals, while preserving SEO value. It scores current content on GEO factors and rewrites weak passages into quotable, citation-length statements. A developer or writer uses it to improve AI-citation odds for existing or new content.16installs109Legal Coglegal-cog is a Claude Code skill that generates legal documents using the CellCog AI service. It drafts contracts, NDAs, terms of service, privacy policies, and legal research memos, and produces PDF output. It depends on the separate cellcog skill for SDK setup and API calls. A developer uses it to draft agreements and policies from a plain-language brief, then has them reviewed by an attorney.16installs110Programmatic SeoThis skill designs and evaluates programmatic SEO strategies for generating search-driven pages at scale from templates and structured data. A developer uses it to decide whether programmatic SEO should be done at all and to plan directory, location, comparison, or integration pages without creating thin or doorway content. It computes a 0-100 Feasibility Index and only proceeds to page-system design when the score clears its threshold.16installs111Slidev Multi Agentslidev-multi-agent creates, edits, themes, builds, and exports Slidev presentations using a script-first workflow with bundled reference docs. A developer uses it when authoring or customizing Slidev decks, themes, or exports from an agent. It provides deterministic scripts and local references so slide work runs the same across Codex, Claude Code, and OpenClaw.16installs112Unum Stratunum-strat designs, audits, and pressure-tests systematic trading strategies for crypto, stocks, ETFs, forex, and futures before capital is exposed. A trader uses it to build a fee-aware plan, audit an existing bot, or check whether a strategy can survive its own costs and slippage. It runs a structured intake of venue, exact costs, and sizing, then applies a fee-hurdle test and capital-bucket routing, defaulting to conservative spot-style execution.16installs113Clankerclanker is a Claude Code skill for deploying ERC20 tokens using the Clanker TypeScript SDK, which creates tokens with Uniswap V4 liquidity pools in a single transaction. A developer uses it to deploy tokens or memecoins on Base, Ethereum, Arbitrum, and other EVM chains, and to configure vesting, airdrops, reward distribution, market cap, and sniper protection. It requires a wallet private key and covers claiming trading-fee rewards.15installs114Content Repurposingcontent-repurposing is a Claude Code skill that turns one piece of content into many formats. A marketer uses it to atomize a blog post, podcast, or video into threads, carousels, newsletters, short-form videos, and quote cards. It provides format-specific conversion recipes and generates media via the inference.sh CLI.15installs115Kpi Dashboard Designkpi-dashboard-design is a skill with patterns for designing effective KPI dashboards. It covers selecting meaningful metrics (SMART KPIs), a strategic/tactical/operational framework, department-specific KPI catalogs for sales, marketing, product and finance, and dashboard layout patterns. It is guidance-oriented rather than code-generating. A developer or analyst uses it when building or improving business dashboards and choosing which metrics to show.15installs116Scrum Masterscrum-master is a data-driven agile coaching skill for Scrum teams. It ships Python tools to analyze sprint velocity, score sprint health across six dimensions, and mine retrospective patterns, plus templates and frameworks for team dynamics and psychological safety. A scrum master or team lead uses it to forecast delivery and diagnose team problems from sprint JSON data. It expects a defined sprint-data JSON schema as input.15installs117Seo Geoseo-geo is a skill for optimizing websites for both traditional search (Google, Bing) and AI search engines (ChatGPT, Perplexity, Gemini, Claude). It runs technical SEO audits, verifies AI-bot access in robots.txt, generates JSON-LD schema, applies the 9 Princeton GEO methods, and does keyword and competitor research through DataForSEO. A developer or marketer uses it to improve search visibility and AI citation. Free scripts work without an API; deeper research needs DataForSEO credentials.15installs118Startup Business Analyst Business Casestartup-business-analyst-business-case generates an investor-ready business case document covering market opportunity, solution, competitive landscape, financial projections, team, risks, and funding ask. A founder uses it during fundraising or strategic planning to turn company inputs into a structured document. It walks through gathering context and assembling each section of the case.15installs119Trade Prediction Marketstrade-prediction-markets builds and tests Polymarket prediction-market trading strategies for YES/NO token trading. A trader or researcher uses it to browse markets, analyze price history, generate strategy code, and backtest against historical data. It exposes six workbench MCP tools and is currently simulation-only with live deployment noted as upcoming.15installs120Content Extractcontent-extract is a Claude Code skill that converts a URL into readable Markdown for downstream OpenClaw workflows. A developer uses it when web_fetch or the browser is blocked or messy, especially for WeChat pages. It probes cheaply first, falls back to the MinerU API, and returns a standardized result contract with source links.14installs121Employment Contract Templatesemployment-contract-templates is a Claude Code skill that helps draft employment documentation such as contracts, offer letters, and HR policies. It walks through confirming jurisdiction, tailoring role-specific terms, and adding confidentiality and IP assignment clauses. A developer or operator uses it to standardize employment paperwork. The skill states its templates are not legal advice and should be reviewed by qualified counsel.14installs122Hft Quant Experthft-quant-expert is a lightweight skill that supplies quantitative-trading expertise for DeFi and crypto derivatives. It walks the agent through defining an entry signal, sizing positions with a fractional Kelly criterion, validating backtests against common biases, and accounting for gas and slippage costs. It includes quick Python formulas for z-score, Sharpe ratio, Kelly fraction and mean-reversion half-life. A developer uses it as a reference when building trading strategies and risk management.14installs123Pci Compliancepci-compliance is a Claude Code skill for implementing PCI DSS compliance when handling payment card data. It covers the 12 core requirements, the four compliance levels, prohibited-versus-storable data, and tokenization and encryption patterns using Stripe. A developer uses it when building or auditing payment flows to reduce PCI scope and avoid storing card data unsafely.14installs124Portfolio ManagerThis skill connects to the Alpaca MCP Server to fetch a brokerage account's live positions and equity, then analyzes asset allocation, diversification, risk metrics, and individual holdings. A developer or investor uses it to review a portfolio and get rebalancing recommendations based on actual current positions. It enriches positions with market and fundamental data and outputs a structured portfolio report.14installs125Senior Pmsenior-pm is an enterprise project-management skill for software, SaaS, and digital-transformation portfolios. It ships Python tools for multi-dimensional portfolio health scoring, quantitative risk analysis (EMV, Monte Carlo, risk matrix), and resource-capacity planning, plus prioritization models like WSJF, RICE, ICE, and MoSCoW. A PM uses it to produce executive reports, prioritize a portfolio, and quantify project risk. It reads a project-data JSON file as input.14installs126Antislopantislop is a Claude Code skill that detects and fixes AI-generated writing patterns in prose. A writer runs it to audit a draft, calculate a slop score, and apply editor-mode rewrites that remove tells like 'delve', 'game-changer', and negative parallelisms. It combines Wikipedia's signs-of-AI-writing patterns with structural detection and before/after reporting.13installs127Coding Agentcoding-agent is a Claude Code skill for delegating coding tasks to other coding agents such as Codex, Claude Code, or Pi through background bash processes. A developer uses it to build features, review PRs in a temp directory, or refactor large codebases by spawning an agent in a target working directory and monitoring it. It documents the PTY requirements per CLI, background session polling, and stdin control.13installs128Coding Routercoding-router is a Claude Code skill that acts as a compatibility entry point for plan-first coding work, routing requests to plan-issue or coding-agent sibling skills. A developer uses it so planning, plan review, and PR review requests dispatch to the right behavior, and so implementation always produces a plan and waits for an explicit APPROVE before any writes. It defines command aliases, a run-event status contract, and non-negotiable approval gates.13installs129Collaborationcollaboration is a Claude Code skill documenting how to collaborate on GitHub projects. It covers working with write versus read-only access, creating forks and branches, opening PRs toward main, doing inline code reviews, addressing review comments, and managing issues on the original repository. A developer uses it when contributing to projects, creating PRs, or reviewing code on GitHub.13installs130Competitor Teardowncompetitor-teardown is a Claude Code skill that runs structured competitive analysis of rival products. A developer or founder uses it for market research, competitive intelligence, or sales enablement, gathering data via the inference.sh CLI and web screenshots. It outputs feature matrices, pricing comparisons, SWOT analyses, and positioning maps.13installs131Contract Reviewercontract-reviewer is a Claude Code skill that reviews business contracts for risks, missing clauses, and unfavorable terms. A user pastes a contract (NDA, MSA, SaaS, SOW, or vendor agreement) and gets a structured risk report. It scores red-flag clauses, checks for missing protective clauses, and recommends whether to sign, negotiate, or walk away.13installs132First Principles Thinkingfirst-principles-thinking is a Claude Code skill that acts as a Socratic coach for reasoning problems down to fundamental truths. It runs a five-step process of stating the problem, surfacing assumptions, challenging each one, finding fundamentals, and rebuilding solutions. A developer uses it when stuck on a problem, making a high-stakes decision, or building something new. It asks one question at a time and ends with a structured summary document.13installs133Last30dayslast30days is a Claude Code skill that researches any topic from the last 30 days across Reddit, X, YouTube, Hacker News, Polymarket, and the web. It parses the user's intent into a topic, target tool, and query type, then runs a research script and surfaces what people are currently discussing, recommending, and debating. It ends by writing copy-paste-ready prompts for a target tool. A developer uses it to get up to speed on a fast-moving topic before writing or building.13installs134Legal Advisorlegal-advisor is a Claude Code skill that drafts privacy policies, terms of service, disclaimers, and other legal notices. It targets technology-law and privacy compliance, generating GDPR, CCPA, LGPD, and COPPA-aware texts with jurisdiction-specific variations and placeholder sections. A developer uses it to produce first-draft legal documents for a SaaS or e-commerce product. It always flags that the output is a template needing attorney review.13installs135Churn Preventionchurn-prevention is a Claude Code skill for reducing SaaS churn, both voluntary cancellations and involuntary failed-payment churn. A developer or founder uses it to design or optimize a cancel flow, build save offers, run exit surveys, and set up dunning email sequences and retry logic. It includes a 5-stage cancel flow, offer-to-reason mappings, a decline-code taxonomy, and a churn-impact calculator script.12installs136Customer Success ManagerThis Claude Code skill scores SaaS customer health, predicts churn risk, and identifies account expansion opportunities. It runs three deterministic Python CLI tools that take a JSON file of customer data and output text or JSON. A customer-success or account team uses it to triage at-risk accounts and prioritize expansion by effort versus impact.12installs137Financial Calculatorfinancial-calculator is a Claude Code skill for financial math including future value, present value, discount and markup pricing, and compound interest. It provides a CLI, a Python API, and an interactive Flask web UI, plus comparison tables across rates and time periods. A developer or operator uses it to project investment growth, price products, or compare loan and discount scenarios. Formula details are documented in a bundled references file.12installs138Full Cycle Developerfull-cycle-developer is a multi-agent orchestration skill that runs a full development cycle silently: code, test, review, fix, then push a PR. The main session spawns a developer subagent, then four parallel review roles (developer, architect, tester, security), aggregates their blocking findings, runs a fix subagent, and opens the final pull request. Prompts are stack-specific (dotnet, rust, python, go). A developer uses it to automate an end-to-end coding task down to a single reviewed PR.12installs139Create Viral Contentcreate-viral-content is a Claude Code skill that turns drafts into high-engagement social content. A developer or marketer uses it when writing posts for Reddit, Twitter/X, LinkedIn, YouTube, or TikTok that need hooks and shareability. It applies a refinement loop, platform templates, and an AI-tell detector backed by cited research.11installs140Inference Optimizerinference-optimizer is an OpenClaw skill that audits token usage and reduces inference cost and latency. On /optimize or /audit it runs an audit script that reports workspace sizes, session counts and next steps, and on approval it purges stale sessions and daily memory stubs. It exists because loaded workspace files and accumulated session history slow the model without an obvious cause. A developer uses it to keep an agent's context lean and fast.11installs141Mineru Extractmineru-extract is a Claude Code skill that uses the official MinerU parsing API to convert a URL, PDF, Office file, or image into clean Markdown plus structured outputs. It submits the source to MinerU, polls for completion, downloads the result zip, and extracts the main Markdown, handling layout, tables, formulas, and OCR. A developer uses it when web_fetch or a browser produces messy content and higher-fidelity parsing is needed. It requires a MINERU_TOKEN.11installs142Nano Banana Pronano-banana-pro is a Claude Code skill that generates and edits images using Google's Nano Banana Pro (Gemini 3 Pro Image) API. It optimizes the prompt first, drafts at low resolution, then renders a final at up to 4K, and can edit an existing image via --input-image. A developer uses it to create or modify images from the command line with automatic provider fallback.11installs143X Articlesx-articles is a Claude Code skill for writing and publishing long-form X (Twitter) Articles. It supplies hook patterns, a section-by-section article template, three named writing styles, and rules for the Draft.js editor. A developer uses it to draft engagement-oriented content and publish it with browser automation over CDP. It also includes a script that reformats Markdown into X-friendly plain text.11installs144Find Skillsfind-skills is a Claude Code skill that helps discover and install agent skills from the open agent skills ecosystem. It uses the npx skills CLI to search skills.sh, present matching options, and install packages from GitHub. A developer uses it when they want to extend the agent's capabilities and suspect a ready-made skill exists. When no skill matches, it suggests scaffolding a new one with npx skills init.10installs145Model Hierarchymodel-hierarchy is a Claude Code skill that cost-optimizes agent operations by routing each task to the cheapest capable model. It classifies tasks as routine, moderate, or complex and maps them to cheap, mid, or premium model tiers, with escalation when a cheaper model already failed and a vision override for image tasks. A developer uses it when deciding which model to use, spawning sub-agents, or cutting cost. It documents a decision algorithm and per-tier model tables.10installs146Notionnotion is a Claude Code skill that documents how to call the Notion API to create, read and update pages, data sources (databases) and blocks. It provides setup steps, the required Notion-Version header, property-type formats, and the 2025-09-03 database-to-data-source changes. A developer uses it to script Notion operations with curl.10installs147Api Gatewayapi-gateway is a Claude Code skill that lets an agent call 100+ third-party APIs through Maton's managed-OAuth passthrough proxy. A developer uses it to hit native endpoints (for example Slack chat.postMessage or Gmail messages) via a single base URL and API key, with Maton injecting the right OAuth token. It bundles routing references documenting each service's path patterns and manages connections through a separate control API.9installs148Brand Identitybrand-identity is a Claude Code skill that builds a complete brand identity for a solo business, from strategy through execution. It defines brand purpose, three values, personality, voice and tone, then a visual system of colors, typography, logo direction, and imagery style, ending in a brand guidelines document. A developer or founder uses it when creating a new brand, rebranding, or making brand decisions consistent.9installs149Brave Searchbrave-search is a Claude Code skill for headless web search and content extraction using the Brave Search API, with no browser required. A developer uses it to search for documentation or facts and to fetch a URL's readable content as markdown. It runs bundled search.js and content.js scripts and needs a BRAVE_API_KEY.9installs150Goggog is a skill that wraps the gog Google Workspace CLI for Gmail, Calendar, Drive, Contacts, Sheets and Docs. It provides ready command patterns for searching and sending email, reading and updating Sheets, exporting Docs, listing Contacts and querying Calendar. It requires a one-time OAuth credential setup and prefers JSON output for scripting. A developer uses it to automate Google Workspace tasks from an agent.9installs151N8n Code Javascriptn8n-code-javascript is a Claude Code skill that guides writing JavaScript inside n8n Code nodes. It covers the two execution modes, data-access patterns with $input, $json and $node, the required array-of-json return format, and common errors like webhook payloads nested under .body. A developer uses it when building or debugging JavaScript logic in an n8n workflow.7installs152N8n Mcp Tools Expertn8n-mcp-tools-expert is a Claude Code skill that documents how to use the n8n-mcp MCP server tools to build workflows. It covers node discovery, configuration validation, workflow management, the two required nodeType formats, and common tool-selection mistakes. A developer uses it when driving n8n-mcp tools to search nodes, validate configs, and create or edit workflows.7installs