Agent skills you can check before you install
Public SKILL.md skills from GitHub repositories with 50+ stars and from the vendors themselves, scanned and scored. Install any of them with one command in Claude Code, Cursor, Codex, GitHub Copilot or Gemini CLI.
45,604 skills from 2,205 repositories · updated 3 hours ago
9,546 skills
- ar-vr-xrtravisjneuman/.claudeAR/VR/XR development with Unity XR, WebXR, ARKit, ARCore, Meta Quest SDK, and spatial computing. Use when building augmented reality, virtual reality, mixed reality applications, or spatial experiences.OtherNo warnings★ 106↓ —
- autonomous_buildingdeeleeramone/PyWryAutonomous PyWry application building using LLM sampling, elicitation, and progress reporting.OtherNo warnings★ 99↓ —
- reading-invoicekazukinagata/shinkoku請求書の画像を読み取り構造化データを返す。 他のスキルから呼び出されるほか、直接ユーザーが呼び出すことも可能。OtherNo warnings★ 366↓ —
- reading-withholdingkazukinagata/shinkoku源泉徴収票の画像を読み取り構造化データを返す。 他のスキルから呼び出されるほか、直接ユーザーが呼び出すことも可能。OtherNo warnings★ 366↓ —
- reading-receiptkazukinagata/shinkokuレシート・領収書・ふるさと納税受領証明書の画像を読み取り構造化データを返す。 他のスキルから呼び出されるほか、直接ユーザーが呼び出すことも可能。OtherNo warnings★ 366↓ —
- reading-deduction-certkazukinagata/shinkoku控除証明書（生命保険料・地震保険料等）の画像を読み取り構造化データを返す。 他のスキルから呼び出されるほか、直接ユーザーが呼び出すことも可能。OtherNo warnings★ 366↓ —
- setupkazukinagata/shinkokuThis skill should be used when the user wants to set up shinkoku for the first time, initialize their configuration, or update their existing settings. Trigger phrases include: "セットアップ", "初期設定", "設定ファイルを作る", "shinkokuの設定", "config", "setup", "始め方", "使い方", "設定を更新", "configを作り直す".OtherNo warnings★ 366↓ —
- reading-payment-statementkazukinagata/shinkoku支払調書の画像を読み取り構造化データを返す。 他のスキルから呼び出されるほか、直接ユーザーが呼び出すことも可能。OtherNo warnings★ 366↓ —
- capabilitieskazukinagata/shinkokuDisplay shinkoku's current capabilities, supported personas, and known limitations. Use when the user asks "what can you do?", "what's supported?", or similar questions.OtherNo warnings★ 366↓ —
- monte-carlo-proactive-monitoringmonte-carlo-data/mc-agent-toolkitGuide users from coverage analysis to monitor creation. USE WHEN user asks what should I monitor, where are my gaps, improve coverage, or wants a systematic approach to monitoring across their data estate.OtherNo warnings★ 93↓ —
- d3-data-tablecxcscmu/SkillLearnBenchHow to build an interactive HTML data table driven by D3.js v6 with sortable columns, row highlighting, cross-component selection, and formatted values. Use this skill when building tabular displays alongside D3 charts, especially when the table needs to stay in sync with chart interactions.OtherNo warnings★ 91↓ —
- pm-poolkonglong87/superPMUse when: 有新需求需要录入跟踪、需求状态需要变更、需要统一管理需求池、需求数量增多需要结构化 Do NOT use when: 需求极少可直接管理、仅需一次性分析无需持续跟踪OtherNo warnings★ 83↓ —
- monte-carlo-troubleshoot-agent-tracesmonte-carlo-data/mc-agent-toolkitTroubleshoots Monte Carlo AI agent alerts and traces — eval score drops, latency/token spikes, trajectory and validation breaches. Not for data incidents (monte-carlo-analyze-root-cause) or monitor creation (monte-carlo-monitoring-advisor).OtherNo warnings★ 93↓ —
- logical-thinkingTheGoat395/Codex-SkillsDetermine what premises and evidence warrant.OtherNo warnings★ 133↓ —
- glm-runnercxcscmu/SkillLearnBenchManage GLM (General Lake Model) configuration and execution. Use this skill when modifying 'glm3.nml' or running the GLM binary to ensure simulation parameters are correctly set and the model runs successfully.OtherNo warnings★ 91↓ —
- maven-build-skipcxcscmu/SkillLearnBenchGuide on how to rebuild Druid using Maven while skipping unnecessary checks. Use this skill when asked to compile or build Druid.OtherNo warnings★ 91↓ —
- mcp-toolsmrpulor-gh/nuphusNuphus — 本地优先的 AI Agent：真实桌面执行力 + 手机第二块屏幕。Local-first AI agent with real desktop execution and dual-device real-time sync.OtherNo warnings★ 89↓ —
- core-data-expertAvdLee/Core-Data-Agent-SkillExpert Core Data guidance (iOS/macOS): stack setup, fetch requests & NSFetchedResultsController, saving/merge conflicts, threading & Swift Concurrency, batch operations & persistent history, migrations, performance, and NSPersistentCloudKitContainer/CloudKit sync.OtherNo warnings★ 318↓ —
- huggingface-trlcxcscmu/SkillLearnBenchUsing Hugging Face TRL library for building custom trainers for Preference Optimization methods like DPO, SimPO, etc.OtherNo warnings★ 91↓ —
- glm-calibrationcxcscmu/SkillLearnBenchCalibrating the General Lake Model (GLM) by modifying parameter bounds for lake simulation.OtherNo warnings★ 91↓ —
- run2_simpo-losscxcscmu/SkillLearnBenchComplete implementation of the SimPO (NeurIPS 2024) loss function with all variants and edge cases - use when reproducing SimPO or implementing reference-free preference optimization.OtherNo warnings★ 91↓ —
- run3_skill-1cxcscmu/SkillLearnBench[COLM'26] SkillLearnBench is the first benchmark for evaluating continual learning methods that automatically generate agent skills.OtherNo warnings★ 91↓ —
- directory-build-organizationdotnet/skillsUSE ONLY for (1) two or more projects with repeated MSBuild policy, targets, or package versions, or (2) an existing Directory.Build.props/targets/rsp hierarchy with an import, placement, or layering defect. A one-project repo without an existing Directory.Build file is outside scope; do not load this skill. Covers centralizing shared policy and targets, Directory.Packages.props across projects, parent hierarchy/import structure, props-versus-targets placement, and folder/project exceptions. An existing hierarchy defect remains in scope when only one project is affected. For property value or condition semantics use property-patterns. Exclude legacy-to-SDK migration and non-MSBuild systems.OtherReview first★ 5.7K↓ —
- evidence-first-engineeringNocodeMrLi/mini-program-engineering-skill-suiteApply evidence-first engineering discipline to any agent-built software project: fact discovery before action, evidence-calibrated status reporting, explicit change boundaries and stage gates, controlled confirmation for risky or external actions, sensitive-information redaction for anything shared publicly, and resumable continuity after interruptions. Use when users ask an agent to take over an existing codebase, deliver a feature across stages, verify whether work is actually complete, judge release readiness, or keep conclusions honest about what is proven versus assumed. This foundation skill is domain-neutral; vertical suites (for example mini-program engineering) build on it by adding domain facts and platform rules.OtherNo warnings★ 91↓ —