Find the perfect skill
for your AI agent
The premium marketplace for AI agent skills and tools.Verified, secure, one-command install.
claude mcp add agentpowers -- npx -y @agentpowers/mcp-server✓ MCP server addednpx @agentpowers/cli search web-scraper✓ Found 3 skills matching "web-scraper"npx @agentpowers/cli install web-scraper✓ Installed to ~/.claude/skills/web-scraperInstall in 30 seconds
One command to install. No config files, no build steps, no restarts. Skills are live in your next session and ready to use in seconds.
Learn how it worksSecurity scanned
Every skill runs through an 8-layer automated pipeline before it reaches you. Malware detection, dependency audit, and AI review.
View security detailsNo coding required
Just chat with your AI assistant to discover, purchase, and install skills. One-click purchase, natural language install - no terminal experience needed.
See how it worksFeatured Skills
Small Business Finance Ops
Six agent skills for small-business money tasks: a router skill plus bill calendar, cash-flow forecast, expense categorizer, month-end close and subscription audit. Works with Claude Code, Codex and Cursor. One-time purchase, no subscription.
KQL Investigation Planner
Turn an authorized defensive security question into a bounded investigation plan and read-only KQL drafts. Receive investigation scope, schema assumptions, staged query drafts, interpretation guidance, and evidence gaps. Provide a sanitized security question, time range, and relevant schema. This advisory workflow does not connect to systems or run queries; validate proposed queries in your authorized environment. One-time purchase; your AI assistant and source-system licenses are separate.
Business Signals to Action Plan
Turn supplied business observations and metrics into a prioritized, evidence-linked action plan. This advisory skill works from the evidence you provide. It produces structured findings and validation guidance. No live connection, execution, deployment, or guaranteed outcome is included.
MVP Scope & Release Readiness
Turn a product brief and delivery evidence into a focused MVP scope and a clearly qualified release recommendation. This advisory skill works from the evidence you provide. It produces structured findings and validation guidance. No live connection, execution, deployment, or guaranteed outcome is included.
Requirements & Acceptance Review
Review requirements, acceptance criteria, and test evidence to find ambiguity and unsupported readiness claims. This advisory skill works from the evidence you provide. It produces structured findings and validation guidance. No live connection, execution, deployment, or guaranteed outcome is included.
Incident Triage & Postmortem
Turn incident notes, logs, and metrics into a sourced triage brief or a practical postmortem. This advisory skill works from the evidence you provide. It produces structured findings and validation guidance. No live connection, execution, deployment, or guaranteed outcome is included.
Dataverse Architecture Review
Review a Dataverse design for table grain, ownership, relationship lifecycle, access boundaries, and solution delivery. This advisory skill works from the evidence you provide. It produces structured findings and validation guidance. No live connection, execution, deployment, or guaranteed outcome is included.
Power Query M Code Review
Review Power Query M transformations for join multiplication, schema drift, conversion errors, and refresh risks. undefined This is a static advisory review based on supplied evidence. No live connection, execution, deployment, or guaranteed outcome is included. Proposed checks require authorized tools and remain unexecuted unless explicitly run.
Three steps. One new skill.
Browse & Discover
Search across seven categories. Filter by skill type, sort by downloads or price, and check security status. Every listing shows what it does, who built it, and whether it passed review.
Purchase & Get License
One-click checkout through Stripe. Instant license code delivery. Check the security report, price, and ratings - full control, always.
Install & Use Instantly
Run one command in your terminal. The skill downloads, installs, and is ready in your next session. No context switching.
Browse by Category
Data
7 skills & agentsDesign
4 skills & agentsDevelopment
24 skills & agentsMarketing
15 skills & agentsProductivity
21 skills & agentsSales
4 skills & agentsSecurity
5 skills & agentsThe numbers speak for themselves
Average Install Time
One command in your terminal. No config files, no dependency conflicts, no setup guides. Purchase to running in under a minute.
Security Scanned
Every skill passes automated vulnerability scanning before it's listed. Code injection, data exfiltration, and access violations are caught automatically.
Creator Revenue Share
We take 15% and handle all payment processing fees. No hidden costs, no surprise deductions. Creators keep exactly 85% of every sale.
Recurring Fees
All skills are one-time purchases. No subscriptions, no hidden charges, no renewal surprises. Buy once, use forever.
Top Creators
Meet the builders behind the most popular skills.
Roy Yuen
william yaohui
Qingwei Hu
Tallyfern
Dani Acosta
rachid hakim
MonRiah Studio
Andy Roberts
Henry Rothschild
Skit E'dizzle
Reiner Weisssieker
idanbeckster
Works with your tools
Connect AgentPowers to your favorite AI platform with dedicated plugins and extensions.
Claude Cowork
Drag & drop install. 24 marketplace tools for Claude Cowork โ search, buy, and install skills from chat.
OpenAI Codex
Full plugin with checkout orchestration, cross-tool install, lifecycle hooks, and bundled skill.
Frequently Asked Questions
What are AI agent skills?
AI agent skills are reusable capabilities that extend what your AI assistant can do. Each skill is a set of instructions, prompts, and workflows packaged for a specific task - like generating cold emails, running SEO audits, or writing unit tests. Once installed, skills integrate directly into any MCP-compatible tool - Claude, Cursor, Windsurf, and more.
How do I install a skill from AgentPowers?
Installing a skill takes about 30 seconds. After purchasing, you receive a license code. Run npx @agentpowers/cli install skill-name --code YOUR-LICENSE-CODE in your terminal. The skill is available immediately in your AI assistant - no configuration or setup required.
Are skills safe to use?
Every skill on AgentPowers goes through automated security scanning before it's listed. Each skill receives a security status: "Verified" (passed all checks), "Warning" (minor findings, disclosed to you), or "Blocked" (rejected). Security findings are visible on every listing so you can make informed decisions.
What's the difference between skills and agents?
Skills are focused, single-task tools - they do one thing well, like generating test suites or auditing accessibility. Agents are more autonomous: they can perform multi-step workflows, use multiple tools, and make decisions on their own. Both install the same way.
How much do skills cost?
Skills range from $5 to $49, with most priced between $8 and $29. All prices are one-time purchases - no subscriptions, no recurring fees.
Can I publish my own skills on AgentPowers?
Yes. Create an account and publish your skill - free or premium. For paid skills, connect Stripe and keep 85% of every sale. AgentPowers takes 15% and absorbs all payment processing fees, far lower than the 30% app stores charge. The minimum listing price is $5.
What payment methods do you accept?
We use Stripe for all payments, which means we accept all major credit and debit cards, Apple Pay, Google Pay, and bank transfers in supported countries. All transactions are encrypted and PCI-compliant.
What if a skill doesn't work as expected?
If a skill doesn't perform as described, contact us at hello@agentpowers.ai. We review all complaints and work with creators to resolve issues for skills that don't deliver on their listed functionality.
What is an AI skill marketplace?
An AI skill marketplace is a platform where developers and creators publish, sell, and distribute reusable capabilities for AI coding assistants. AgentPowers is the first premium AI skill marketplace built on the open Model Context Protocol (MCP) standard, supporting 12+ AI platforms including Claude Code, Cursor, Windsurf, Codex, and Gemini CLI. Every listing is security-scanned through an automated 8-layer pipeline before publication.
What is the Model Context Protocol (MCP)?
The Model Context Protocol (MCP) is an open standard created by Anthropic that defines how AI assistants communicate with external tools and data sources. It works like USB for AI: a universal interface that lets any MCP-compatible platform (Claude, Cursor, Windsurf, GitHub Copilot, and others) use the same skills and agents without platform-specific code.
How does AgentPowers compare to other AI tool platforms?
AgentPowers is the only MCP-first marketplace with automated security scanning, creator payouts (85% revenue share), and cross-platform compatibility across 12+ AI tools. Unlike closed plugin ecosystems, AgentPowers skills work everywhere MCP is supported. Unlike free registries, every listing is verified through an 8-layer security pipeline including static analysis, malware detection, and AI-powered code review.
Ready to supercharge
your AI workflow?
Join AgentPowers and get instant access to verified skills and agents built by the community.