AI Skills Hub - Browse 180,000+ Agent Skills

Hybrid semantic + keyword search across memory files using BM25 for keyword matching and OpenAI embeddings for semantic reranking.

Use when you need to synchronize commands and skills from ~/.agents/ to AI tool directories like Claude or CodeBuddy using symbolic links

Deploy Grafana OSS to Azure Container Apps. Use when deploying Grafana for metrics, logs, and traces visualization with optional PostgreSQL backend.

Writes tests for LangGraph agents and graphs with pytest. Use when testing state machines, graph nodes, agent flows, or partial execution paths.

Enforce an E2E user-story gate for trade_canvas work: during planning require a complete end-to-end user story covering the main flow; during development require passing that E2E story before declaring done; during delivery require reporting the story, flow, produced data, and concrete evidence (commands + outputs/artifacts). Use when planning/implementing any feature or fix that spans multiple components or changes behavior.

Use at the final acceptance stage of trade_canvas work (before claiming 'done' / delivery): requires FE+BE integrated Playwright E2E gate to pass (exit code 0) and requires attaching concrete evidence (commands + outputs + artifacts path).

vibe-commander.config.json 설정 파일의 스키마 설계 및 Zod 4 검증 패턴을 가이드합니다. 설정 파일 구조, 스키마 정의, 설정 검증, 새 설정 필드 추가 요청 시 사용합니다.

Synthesizes speech from text using Doubao (ByteDance) TTS V3 API. Supports 100+ voice types, emotions, speed control, and audio formats (mp3/wav/pcm/ogg_opus). Use when converting text to audio, generating voice content, reading text aloud, or TTS tasks. 豆包语音合成，文字转语音，朗读文本。

Transcribes audio to text using Doubao (ByteDance) ASR API. Supports streaming recognition, WAV/MP3 formats, ITN, punctuation, and Chinese/English speech. Use when converting speech to text, transcribing audio files, or ASR tasks. 豆包语音识别，语音转文字，音频转录。

LP本番ビルド。画像最適化、タグ注入、minifyを実行。ビルドして、公開準備して、と言われたら使用。

Package voxtype for release. Creates deb and rpm packages from binaries. Use when building distribution packages.

Build production-ready RAG chatbots with LangGraph/LangChain, Pydantic schemas, and Supabase vector store using Gemini embeddings. Use when creating conversational AI that answers questions from documents (PDFs, web pages, text files), building document Q&A systems, or implementing retrieval-augmented generation pipelines.

Remove backgrounds from images using FAL.ai's BiRefNet model. Use when users ask to remove background, make transparent PNG, extract subject from image, or create cutouts. Trigger phrases include "remove background", "transparent background", "cut out", "extract subject", or any background removal request.

Run GitHub Actions workflows locally before pushing to validate CI/CD pipelines, catch errors early, and reduce iteration time. Uses [nektos/act](https://github.com/nektos/act) to simulate GitHub Actions runner environment.

透過 CDP 控制 Chrome 瀏覽器執行自動化任務。Use when needing to control browser, scrape websites, take screenshots, interact with web apps, or automate browser tasks. Triggers on "瀏覽器自動化", "截圖網頁", "scrape", "browser automation", "CDP".

用 AI 生成圖片。Claude 生成詳細描述，Puppeteer 自動操作 Gemini 生圖並下載。Use when user says "生成圖片", "generate image", "幫我配圖", "create illustration", or needs images for blog/website.

3. **Opportunity Extraction** - Extracts opportunity details from project pages - Downloads associated documents (PDFs, DOCX) - Stores structured data in SQLite and Supabase

Implement images into page sections. Converts from /public/raw/[page]/ to WebP with responsive variants, replaces CDN placeholders with Next.js Image component. Use after midjourney-prompts skill. Triggers on "implement images", "add images to page", "replace placeholder images".

Use git worktrees to isolate tasks. Never switch branches in base folders.

Blender best practices, Python code snippets, and design-for-3D-printing rules for the Blender MCP server. Use when creating 3D models, manipulating geometry, applying materials, or performing operations like bisecting, boolean operations, and mesh editing. Also use when designing printable parts — covers orientation, edge treatment, hole design, fit features, support avoidance, and dimensional tolerances for FDM/FFF printing. Contains working bpy code patterns discovered through iteration.

Analyze screen recordings and websites to extract implementation specs, design systems, and UI patterns.

Guide for creating Claude Code slash commands. Use when the user wants to create a new slash command, update an existing slash command, or asks about slash command syntax, frontmatter options, or best practices.

Automatically convert arXiv papers to well-structured Markdown documentation. Invoke with an arXiv ID to fetch materials (LaTeX source or PDF), convert to Markdown, and generate implementation-ready reference documentation with preserved mathematics and section structure.

Best practices for searching and interacting with Apple Mail on macOS

Start or check Remotion Studio and return preview URL/PID/log paths after lyric video work, or when user asks to launch development preview. Use for requests like "start Remotion", "open preview", "開発環境を起動して確認", or "リリック動画作成後の確認".

Look up Pokemon Trading Card Game prices from the TCGCSV API. Use this when users ask about card prices, market values, or availability for specific Pokemon cards in English sets. Supports flexible card name matching, rarity abbreviations (IR, SIR, etc.), and "did you mean?" suggestions for misspelled set or card names. Returns current market price, listing prices, and card details without loading entire CSV files into context.

Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.

Enhance Project Zomboid server atmosphere via RCON. Use for broadcasting narrative messages to players, giving items/XP rewards, spawning vehicles, triggering world events (hordes, helicopters, gunshots), and controlling weather. Focus on making the server feel alive with storytelling and dynamic events.

Install, configure, and troubleshoot MCP servers for Claude Code and Cursor. Use when setting up new MCP connections, debugging MCP issues, or selecting which MCP servers to add for a project.

剪映 (JianYing) AI自动化剪辑的高级封装 API (JyWrapper)。提供开箱即用的 Python 接口，支持录屏、素材导入、字幕生成、Web 动效合成及项目导出�?

Analyze Terraform plan JSON output for AzureRM Provider to distinguish between false-positive diffs (order-only changes in Set-type attributes) and actual resource changes. Use when reviewing terraform plan output for Azure resources like Application Gateway, Load Balancer, Firewall, Front Door, NSG, and other resources with Set-type attributes that cause spurious diffs due to internal ordering changes.

视频智能剪辑工具。支持 YouTube、Twitter/X、Instagram、TikTok 和本地视频。下载视频和字幕，AI 分析生成精细章节（几分钟级别）， 用户选择片段后自动剪辑、翻译字幕为中英双语、烧录字幕到视频，并生成总结文案。 使用场景：当用户需要剪辑 YouTube/Twitter/Instagram/TikTok 视频、处理本地视频、生成短视频片段、制作双语字幕版本时。 关键词：视频剪辑、YouTube、Twitter、Instagram、TikTok、本地视频、字幕翻译、双语字幕、视频下载、clip video、剪片

Automated style and structure validation for Odoo documentation. Use when creating or modifying RST files to ensure compliance with Odoo's heading hierarchy, line length limits, image metadata requirements, and UI element formatting rules. This skill provides validation scripts and a comprehensive rule reference for maintaining documentation quality.

Autonomous setup and management of external service monitoring using UptimeRobot (HTTP endpoint monitoring) and Healthchecks.io (heartbeat/Dead Man's Switch monitoring). Use when setting up monitoring for Cloud Run Jobs, VM services, or investigating monitoring configuration issues. Includes Pushover integration, validated API patterns, and dual-service architecture.

Automated workflow for generating AI-powered content using agent_codex.py. Handles prompt setup, batch processing, validation, and output management for InsightfulAffiliate and NextGenCopyAI content.

Image resizing tool optimized for AI API file size limits. Resize images for Gemini, Claude, and other AI APIs.

Server domain expert for code review and validation. Use this skill when reviewing pull requests that touch src/server/, when analyzing server functions, middleware, database queries, or authentication code, when detecting Cloudflare Workers compatibility issues, or when a senior AI agent needs to validate code against Server domain best practices. Provides automated violation detection and actionable PR comments.

Integrations domain expert for code review and validation. Use this skill when reviewing pull requests that touch src/integrations/, when analyzing third-party API client implementations, when validating Vercel AI SDK tool definitions, when detecting domain violations (module-level instances, missing descriptions, debug logging), or when a senior AI agent needs to validate code against Integrations domain best practices. Provides automated violation detection and actionable PR comments.

Types domain expert for code review and validation. Use this skill when reviewing pull requests that touch src/types/, when analyzing type definitions across the codebase, when detecting cross-domain type violations (types exported from wrong domain), or when a senior AI agent needs to validate code against Types domain best practices. Provides automated violation detection and actionable PR comments.

Components domain expert for code review and validation. Use this skill when reviewing pull requests that touch src/components/, when analyzing React component patterns (CVA, Radix UI, shadcn/ui), when detecting inline styling violations (styled elements outside ui/ directory), when validating component structure (forwardRef, data-slot, props), or when a senior AI agent needs to validate code against Components domain best practices. Provides automated violation detection and actionable PR comments.

Client domain expert for code review and validation. Use this skill when reviewing pull requests that touch src/client/, when analyzing TanStack Query hooks, Zustand stores, PGlite database operations, or client actions, when detecting client domain violations (missing query guards, derived state in stores, improper db access), or when a senior AI agent needs to validate code against Client domain best practices. Provides automated violation detection and actionable PR comments.

Activate this skill when the user: - Provides a YouTube URL and wants transcription - Requests transcript extraction, captions, or subtitles - Wants to analyze/summarize YouTube video content

Run tests from skill examples and generate a report (project)

Turborepo monorepo build system guidance for task pipelines, caching, and CI optimization

Defines the standard 8-category taxonomy for classifying project files: Config, Tests, Docs, Scripts, Source Code, Data, AI Tooling, and Other. REFERENCE THIS SKILL whenever categorizing files, auditing project structure, or answering 'what types of files are here.' Apply these categories and patterns whether using the bundled script or manual analysis.

Interact with UniFi Network controllers via the LOCAL Integration API v1 (NOT the cloud Site Manager API). Use when: (1) querying network devices, clients, or statistics, (2) managing networks, WiFi SSIDs, firewall zones, ACL rules, (3) generating/managing hotspot vouchers, (4) listing VPN servers, WAN interfaces, RADIUS profiles. Supports UDM, UDM-Pro, UDM-SE, and Cloud Key controllers with full CRUD operations.

Generate high-quality images using Google's Gemini image generation models (Nano Banana Pro). This skill should be used when users request AI-generated images, text-to-image generation, image creation from descriptions, or product mockups. Supports multiple aspect ratios, resolutions up to 4K, and advanced features like reference images and text rendering. Uses gemini-3-pro-image-preview model by default.

Browser automation, debugging, and performance analysis using pure Chrome DevTools Protocol (CDP). Use for automating browsers, taking screenshots, analyzing performance, monitoring network traffic, web scraping, form automation, and JavaScript debugging without high-level libraries.

提供一個可重用的 skill，用於指導使用者將 SRT 格式字幕轉換為 VTT（WebVTT）格式。此 skill 支援多語言（英文、繁體中文、簡體中文），並示範如何使用現成的 `srt-to-vtt` Python 套件與備選工具進行完整轉換與格式/編碼自動偵測。

專業音訊品質檢測員：分析本地音訊檔案的編碼、取樣、聲道與音量特徵，並依規則給出品質評等與報告。使用情境：當使用者提供本地 .mp3/.wav/.m4a 路徑，需快速生成技術檢測報告與品質建議。

List all sheet names in an Excel workbook. Use this when users need to know what sheets are available in a workbook before processing.

Multi-phase schema optimization workflow orchestrator. Creates session directories, spawns phase agents sequentially, validates outputs, aggregates results. Trigger: "run schema optimization", "optimize schema workflow", "execute schema phases"

Bouffalo SDK开发指南，提供完整的Bouffalo芯片系列开发文档和参考资料。当需要使用bouffalo_sdk开发IoT/MCU应用、了解SDK架构、配置构建系统、使用外设驱动、开发无线功能、管理电源、调试固件时使用此skill。

Orchestrate a configurable, multi-member CLI planning council (Codex, Claude Code, Gemini, OpenCode, or custom) to produce independent implementation plans, anonymize and randomize them, then judge and merge into one final plan. Use when you need a robust, bias-resistant planning workflow, structured JSON outputs, retries, and failure handling across multiple CLI agents.

Fetch and read transcripts from YouTube videos. Use when you need to summarize a video, answer questions about its content, or extract information from it.

Mandatory single-file editing Skill — the ONLY permitted way to modify or create files. Strictly one file per invocation, tmp workspace inside project root (.robust-edit-tmp.XXXXXX, auto-cleaned), basename-only patch headers, saved traceable patch next to target, automatic CRLF normalization + \\n unescape robustness loop.

Trigger when: (1) User mentions "manim" or "Manim Community" or "ManimCE", (2) Code contains `from manim import *`, (3) User runs `manim` CLI commands, (4) Working with Scene, MathTex, Create(), or ManimCE-specific classes. Best practices for Manim Community Edition - the community-maintained Python animation engine. Covers Scene structure, animations, LaTeX/MathTex, 3D with ThreeDScene, camera control, styling, and CLI usage. NOT for ManimGL/3b1b version (which uses `manimlib` imports and `manimgl` CLI).

Helps diagnose and fix failing tests in pytest. Use when tests fail unexpectedly, tests are flaky, test output is confusing, or need to understand test failures. Parses pytest output and suggests fixes.

Persistent email tracking across sessions. Differentiates new emails from seen. Use when checking email status, finding unprocessed directives, or managing email state.

Next.js Metadata API for SEO, OpenGraph tags, structured data, and social sharing. Use when implementing metadata, SEO, or social media previews.

MySQL database administration using a pure TypeScript/JavaScript protocol implementation. Use this when users need to connect to MySQL databases, run queries, manage users and permissions, backup databases, or perform other DBA tasks. No external MySQL client libraries required.

Implements Ably realtime messaging with Pub/Sub patterns, token authentication, presence management, and LiveObjects. Use when building chat, notifications, collaborative apps, or realtime features requiring state synchronization.

Implements Radix UI primitives using architectural patterns (asChild composition, component abstraction), applying best practices (mandatory prop spreading, ref forwarding, portal patterns), and avoiding anti-patterns. Use when building accessible UI components, integrating with design systems, or managing portals.

Implements Flyway database migrations in Spring Boot 3.2.5. Covers single/multi-datasource patterns, multi-schema management, callbacks, best practices (immutability, zero-padding, SQL-first), workarounds (baseline, manual scripts), and anti-patterns. Use for schema versioning and lifecycle management.

ずんだもん＆めたんの掛け合い紹介動画を作成するためのスキルです。

Remotion + Qwen-TTS（MLX）で動画を生成する際の完全ガイド。音声生成、口パク同期、フレーム計算、そして必須の検証フェーズまでカバー。動画生成時に使用。

Pre-push validation checklist (cargo fmt, clippy with zero warnings, feature flag testing, test suite), CI monitoring, merge process, and release quality gates. Use when preparing to push code, validating changes before PR, running CI checks, merging PRs, or preparing releases.

GIGAZINE（ギガジン）のAIカテゴリページから指定期間（デフォルト1週間）の記事を取得し、各記事の要約を含むマークダウンファイルを生成する。使用タイミング：(1) ユーザーがGIGAZINEのAI記事まとめを要求した時、(2) テックニュースの定期的な要約が必要な時、(3) https://gigazine.net/news/C48/ から記事を取得したい時。

Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.

Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.

Persistent memory system for AI agents with semantic, episodic, and procedural memory types. Use when users want to (1) remember facts, preferences, or context across sessions, (2) track interaction history and experiences, (3) store reusable workflows or procedures, (4) build personalized agents that learn from conversations, or (5) implement any form of long-term memory for AI applications.

Automatically manages marketplace catalog updates, syncs marketplace.json, and handles plugin distribution when user mentions marketplace update, sync catalog, or add to marketplace. specific to claude-code-plugins two-catalog system.

Automatically validates claude code plugin structure, schemas, and compliance when user mentions validate plugin, check plugin, or plugin errors. runs comprehensive validation specific to claude-code-plugins repository standards.

Specialized skill for analyzing JavaScript bundle sizes, identifying optimization opportunities, detecting unused code, and monitoring performance budgets. Use when bundle size exceeds limits, analyzing code splitting, identifying heavy dependencies, or optimizing JavaScript payload.

Manage OpenClaw round-robin model selection. Use when user asks about round-robin, types /round-robin or /round-robin edit, or wants to change which models rotate.

Expert system administrator for NixOS remote cache servers on modest hardware. Use when setting up, optimizing, troubleshooting, or planning NixOS cache servers, especially on low-spec machines with limited RAM or CPU. Provides realistic assessments, creative workarounds, hardware upgrade planning, direct network configuration, and honest "this won't work" feedback when needed. Specializes in local/on-prem deployments with resource constraints.

Manage git worktrees with shell scripts for listing, creating, and deleting worktrees. Use when working with git repositories and need to create worktrees with new branches, delete existing worktrees, or view all worktrees. Particularly useful for parallel feature development or testing multiple branches simultaneously.

print(f"Found {len(ready_rows)} repositories to analyze") ```

feedback_data = {} for row in ws3.iter_rows(min_row=2, values_only=True): email_id, grade, category, persona, response, attempts, status = row if status == "Ready" and response: feedback_data[email_id] = { 'grade': grade, 'response': response }

This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.

SvelteKit data flow guidance. Use for load functions, form actions, and server/client data. Covers +page.server.ts vs +page.ts, serialization, fail(), redirect(), error().

Use when creating or developing Claude Code plugins. Scaffolds structure, adds components, sets up development environments, and provides guidance on plugin architecture and best practices.

Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.

Expert guidance for connecting Cloudflare-managed domains to Firebase Hosting, including manual setup steps, DNS configuration, SSL/TLS provisioning, and automated Cloudflare API scripts for DNS record management. Use when configuring custom domains for Firebase-hosted applications with Cloudflare DNS.

Generate comprehensive research reports combining Claude AI analysis with Google search results. Use when the user requests: (1) Research reports on technical questions with web research, (2) Analysis combining AI insights and current web sources, (3) PDF reports synthesizing multiple information sources, (4) Commands like /summarise-claude-google. Produces modern, tech-style PDF reports with Claude's answer, top 10 search results, and synthesized conclusions.

Search Maven artifacts by GAV (GroupId, ArtifactId, Version) using Aliyun Maven repository API with intelligent version sorting

Measure property data quality using completeness (60%) + confidence (40%) formula. Quality tiers are Excellent (95%+), Good (80-94%), Fair (60-79%), Poor (<60%). Track field lineage by source (Assessor 0.95, CSV 0.90, Web 0.75, AI 0.70). Use when calculating data completeness, tracking field provenance, enforcing CI/CD quality gates, or generating data quality reports.

Evaluate buyer kill-switch criteria with weighted severity threshold. HARD criteria (HOA=$0, beds>=4, baths>=2) cause instant fail. SOFT criteria (sewer, garage, lot, year) accumulate severity scores - threshold >=3.0 fails, 1.5-3.0 warns. Use when filtering properties, validating purchases, or checking buyer requirements.

Use this skill when the user wants to build tool/scripts or achieve a task where using data from the Hugging Face API would help. This is especially useful when chaining or combining API calls or the task will be repeated/automated. This Skill creates a reusable script to fetch, enrich or process data.

Create a new MultiQC module from a GitHub issue or feature request. This skill guides you through the complete process of implementing, testing, and submitting a new module for parsing bioinformatics tool output.

Test Chrome extensions with stealth automation on bot-protected sites. Use when testing extensions on X.com, LinkedIn, or other sites with CDP detection. Provides cross-script logging, screenshot capture, and HTML analysis. Handles isolated world communication.

Reset and seed the Secret Santa database for testing. Use when you need to clear all data and start fresh, or populate the database with initial participant data.

Release a new version of atomic-agents to PyPI and GitHub. Use when the user asks to "release", "publish", "deploy", or "bump version" for atomic-agents.

Discord Bot API를 이용해 지정된 채널로 텍스트 메시지를 전송하고, YouTube 워크플로 결과를 실시간으로 공유하는 스킬입니다.

YouTube 영상의 자막/메타데이터를 JSON 형식으로 추출하는 스킬입니다. youtube-transcript-api(권장) 또는 Apify API를 사용합니다. 추출된 결과는 youtube-summarizer 등 후속 스킬의 입력으로 사용됩니다.

YouTube 자막과 메타데이터를 입력받아 교육용으로 활용 가능한 구조화 요약(핵심 주제, 논리 전개, 데이터 인사이트, 탐구 질문 포함)을 Markdown 템플릿으로 만들어 주는 스킬입니다. Claude API를 사용한 자동 요약 생성 지원.

Fetch detailed information about YouTube videos using Google's YouTube Data API v3. This skill should be used when users request information about YouTube videos such as title, channel, views, likes, description, or other metadata. Accepts YouTube URLs or video IDs as input.

Manage and maintain the Obsidian knowledge vault. Process daily notes, tag untagged content, refactor notes, and capture Claude sessions. USE WHEN user says "process daily notes", "tag my notes", "clean up vault", "organize notes", "save this conversation", or needs vault maintenance.

Board-specific implementation for creating and managing issues. Use when: creating issues on ZenHub/Jira/Linear, moving issues between pipelines, adding to sprints, setting board-specific fields. For story templates and estimation, use project-management skill. On first use in a project, run setup if .claude/agile-board-config.json is missing.

Generates comprehensive database documentation including ERD diagrams, data dictionaries, and schema analysis. Creates Mermaid, PlantUML, and DBML diagrams from database schemas. Use when documenting databases, creating ERDs, or generating data dictionaries. Trigger with "document database", "create ERD", "generate data dictionary".