Trending

Current operator watchlist

Top 10 Trending AI Skills & MCP Servers This Week

Weekly snapshot of the fastest-moving AI agent skills and MCP-adjacent projects, ranked by repository momentum, discovery signals and practical workflow fit. MCP registries, workflow automation, browser agents, memory systems, and reusable coding-agent skills are leading this refresh.

Installation checks

Verify the source before you install

Check the official source, supported client, token scope and rollout steps before installing a skill or MCP server.

Package checks

Check the package before copying its install command

Exact package names are easy to find. Before you install, check who publishes the package, what it can access and which clients support it.

Directory checks

What to verify in a directory

Compare directories by source quality, review dates and practical setup guidance. A larger catalog or frequent update banner does not establish that a server fits your needs.

Large directory volume

Risk: Readers may assume a bigger raw index is fresher or more complete.

What to check: Verify the original source, installation requirements and workflow fit of each shortlisted tool.

Frequent refresh claims

Risk: An update banner does not show which package details were actually checked.

What to check: Compare the review date with upstream releases and the current official documentation.

Scenario pages and quality scores

Risk: A general score may not reflect the permissions and failure modes of your workflow.

What to check: Check task-specific setup instructions, permission scope and a rollback plan.

Public research assets

Risk: A report may omit the source, test conditions or date behind a recommendation.

What to check: Look for source links, dates and reproducible evidence behind each recommendation.

Follow-up routes

Turn this trend list into the next implementation page

Choose a guide for browser MCP execution, verified MCP discovery or server comparison, depending on the task you want to complete.

This Week's Leaderboard

#Skill / ServerCategoryPublic SignalTrend Signal
1n8n agent workflowsWorkflow Automationlarge OSS baseAgent skills are shifting from code snippets into repeatable business workflows
2LangflowAgent Builderslarge OSS baseVisual agent builders are competing with code-first frameworks for operator attention
3GitHub MCP ServerMCP Coreregistry signalOfficial and registry-backed MCP servers are becoming default trust anchors
4MCP registriesDiscovery Infrastructuredaily freshness signalStatic MCP lists are being replaced by update-aware registries and scored directories
5Browser-use and Stagehand-style agentsBrowser Agentsfast-moving categoryBrowser automation is moving from QA helper to agent runtime primitive
6Agent memory systemsAgent Memorystrong category demandMemory is still a durable adoption blocker for agent teams
7Claude and Codex skillsCoding Agent Skillsoperator demandSkills are becoming a packaging layer for repeatable coding-agent behavior
8OpenWebUI and Ollama local stacksLocal Agent Runtimebroad OSS baseLocal AI stacks are entering agent-skill discovery paths
9DeepAgents-style research agentsResearch Agentscomparison signalResearch agents support tasks that require multiple sources and follow-up investigation
10Repository knowledge systemsRepository Knowledgesteady category demandRepository knowledge tools help agents navigate and explain codebases

Weekly Analysis

Fresh listings still need a source check. Confirm who publishes a package, which permissions it requests and whether its current documentation supports your client before installing.

Use registries to find the original source. Compare registry entries with provider documentation and repository activity, then review installation risks and permission scope.

Start with the workflow you need. Choose between browser automation, workflow automation, agent memory, local runtimes, research and repository knowledge, then compare suitable tools.

Category Breakdown

Workflow Automation

1 in top 10

Compare workflow triggers, credentials and rollback options before choosing an automation tool

MCP Core

2 in top 10

Strong - official registry, source verification, and scored directories are becoming default discovery expectations

Browser Agents

1 in top 10

Rising - browser control needs replay, screenshot, credential, and permission boundaries

Agent Memory

1 in top 10

Durable - memory remains a recurring adoption blocker for team-grade agents

Coding Agent Skills

1 in top 10

Growing - reusable skills are becoming a packaging layer for Claude, Codex, Cursor, and similar tools

Local Agent Runtime

1 in top 10

Growing - local stacks affect whether sensitive workflows can be adopted

Research and Repository Knowledge

2 in top 10

Compare source quality, repository access and the evidence returned by each tool

Why Each Skill Is Trending

#1

n8n agent workflows

large OSS base

Define triggers, credentials and rollback steps when using n8n-style agent automations.

#2

Langflow

large OSS base

Langflow keeps appearing in broad AI-agent directories because teams want a visible builder layer before they standardize on a deeper runtime.

#3

GitHub MCP Server

registry signal

Check the official GitHub MCP source, supported installation method and permissions before connecting a repository.

#4

MCP registries

daily freshness signal

PulseMCP, Smithery, the official registry, and fast-refresh directories all reinforce that discovery freshness is now a product feature, not a blog afterthought.

#5

Browser-use and Stagehand-style agents

fast-moving category

Use screenshots, replay and clear permission boundaries to review browser actions before automating a workflow.

Frequently Asked Questions

How is the weekly trending list calculated?

The list considers repository signals, category fit and practical workflow use. Compare those signals with official documentation and your requirements before choosing a tool.

How often is this page updated?

The trending list is updated weekly, typically on Mondays. Each update reflects the previous week's growth data. Historical snapshots are preserved for trend analysis.

What does stars growth mean?

Repository signals provide context. Compare them with workflow fit, maintenance activity and installation guidance before choosing a tool.

Can I submit a skill for the trending list?

The list is data-driven, not curated by submissions. However, if your skill or MCP server has strong growth signals, it will naturally appear. Submit your project to our main directory for broader visibility.

Why do browser, memory, and repository tools dominate this week?

These categories address common workflow needs: verified MCP sources, inspectable browser actions, persistent memory, source-code context and reusable instructions.

What is the difference between trending and best?

Trending measures momentum — which tools are growing fastest right now. Best (see our Best MCP Servers guide) measures overall quality and utility regardless of current growth. A mature, excellent tool may not be trending if adoption is already saturated.

How can I track trends over time?

Follow project releases and activity on GitHub, and compare dated snapshots when available. Keep notes on the tools relevant to your workflow.

Which pages should I open after this weekly trend snapshot?

Open the MCP browser automation page for browser-agent signals, the MCP Apps directory for source verification, and the Best MCP Servers guide for install-ready server shortlists.

Related Resources

AE
AgentSkillsHub Editorial TeamAI Agent Infrastructure Reviewers

The AgentSkillsHub editorial team evaluates MCP servers, Claude skills, and AI agent integrations for security, reliability, and practical deployment readiness. Every listing undergoes permission audit, README analysis, and operational risk triage before publication.

  • Reviewed 450+ MCP server repositories
  • Developed security grading methodology (A-F)
  • Published agent deployment safety guidelines
Updated: 2026-05-30github