Trending

Week of 2026-05-25 to 2026-05-30

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, competitor freshness patterns, weekly discovery signals, GSC demand, and practical implementation fit. MCP registries, workflow automation, browser agents, memory systems, and reusable coding-agent skills are leading this refresh.

May 30 pulse

Today's trend response should be source-first, not volume-first

Competitors can refresh huge MCP lists quickly. AgentSkillsHub should answer with pages that make installation safer: official source checks, client handoff notes, token scope, and workflow rollout guidance.

Database MCP setup is now a trust query

Prioritize Neon CLI and database MCP pages that explain source authority, token review, and safe branch-first workflows.

Fast directories win on freshness

Counter with visible refresh dates, practical runbooks, and source-validation lanes instead of trying to match raw listing count.

Security research is moving toward MCP exploitability

Promote pages that separate read-only resources from tools that can mutate files, browsers, databases, or SaaS accounts.

Browser and workflow agents remain high-intent

Keep browser automation, n8n, and memory pages tied to screenshots, permissions, replay, and rollback controls.

MCPB and MCP Apps split extension intent

Use desktop extension pages for local-server packaging and MCP Apps pages for interactive UI workflows.

Competitor response

What fast-refresh directories changed in this market

The agent-skills market is no longer just a static directory race. Fast-refresh competitors now emphasize raw index size, scenario pages, quality scoring, and public research assets. AgentSkillsHub should answer with stronger trust signals, practical implementation pages, and visible refresh cadence.

Large directory volume

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

Our response: Keep our directory smaller but explain source trust, install risk, and practical rollout pages better.

Frequent refresh claims

Risk: An 8-hour update banner can make a static weekly page look stale.

Our response: Show explicit weekly pulse dates, next refresh targets, and pages being updated from the pulse.

Scenario pages and quality scores

Risk: Scenario coverage captures high-intent users who do not know a tool name yet.

Our response: Route trend demand into scenarios, MCP apps, security guide, and enterprise procurement pages.

Public research assets

Risk: Blue-book style reports create authority beyond simple listings.

Our response: Turn weekly trend scans into evidence-backed reports, not just a changing top-10 table.

May 2026 follow-up routes

Turn this trend list into the next implementation page

This refresh connects the weekly signal feed to the pages that should absorb search demand next: browser MCP execution, verified MCP discovery, and server shortlist comparison.

Browser MCP execution

Stagehand and Agent Browser signals point to browser-control workflows that need screenshots, replay, and safety boundaries.

Verified MCP discovery

MCP routing demand should land in a source-aware directory before users copy install commands from fast-refresh but low-context lists.

Server shortlist refresh

Database, search, browser, and DevOps servers should be compared as workflow lanes, not as one flat package list.

This Week's Leaderboard

#	Skill / Server	Category	Public Signal	Trend Signal
1	n8n agent workflows	Workflow Automation	large OSS base	Agent skills are shifting from code snippets into repeatable business workflows
2	Langflow	Agent Builders	large OSS base	Visual agent builders are competing with code-first frameworks for operator attention
3	GitHub MCP Server	MCP Core	registry signal	Official and registry-backed MCP servers are becoming default trust anchors
4	MCP registries	Discovery Infrastructure	daily freshness signal	Static MCP lists are being replaced by update-aware registries and scored directories
5	Browser-use and Stagehand-style agents	Browser Agents	fast-moving category	Browser automation is moving from QA helper to agent runtime primitive
6	Agent memory systems	Agent Memory	strong category demand	Memory is still a durable adoption blocker for agent teams
7	Claude and Codex skills	Coding Agent Skills	operator demand	Skills are becoming a packaging layer for repeatable coding-agent behavior
8	OpenWebUI and Ollama local stacks	Local Agent Runtime	broad OSS base	Local AI stacks are entering agent-skill discovery paths
9	DeepAgents-style research agents	Research Agents	comparison signal	Long-horizon research agents are becoming comparison-page traffic magnets
10	Repository knowledge systems	Repository Knowledge	steady category demand	Repo-to-wiki and code-context systems remain high-value support pages

Weekly Analysis

Directory freshness has become the visible battleground. Competitors are no longer presenting agent skills as a small curated list. They now emphasize large indexes, frequent updates, scenario pages, and public research artifacts. Our page needs to show why a security-first shortlist is safer than copying the newest package from a raw feed.

MCP discovery is moving toward registry-backed trust. Official registry work, PulseMCP, Smithery, and similar catalogs make source verification part of the product. AgentSkillsHub should keep pairing discovery with install risk, permission scope, maintenance signals, and team rollout guidance.

Scenario pages should absorb the next update batch. Browser automation, workflow automation, agent memory, local runtimes, research agents, and repository knowledge are the clusters where users know the problem before they know the tool name.

Category Breakdown

Workflow Automation

1 in top 10

Rising - scenario-rich workflow pages are now a stronger competitor signal than flat tool lists

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

Growing - repo and research agents attract evaluator queries that need evidence-first coverage

Why Each Skill Is Trending

n8n agent workflows

large OSS base

Competitor directories are emphasizing workflow volume and scenario coverage, so n8n-style agent automations need implementation pages with triggers, credentials, and rollback checks.

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.

GitHub MCP Server

registry signal

The MCP ecosystem is moving toward registry and source-verification flows, so GitHub MCP deserves current install, permission, and team-policy coverage.

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.

Browser-use and Stagehand-style agents

fast-moving category

Teams need browser actions with screenshots, replay, and safety boundaries. That demand maps directly to our browser automation scenario coverage.

Frequently Asked Questions

How is the weekly trending list calculated?

We combine weekly discovery emails, repository momentum, category fit, GSC demand, and editorial usefulness. The list prioritizes tools that map to real implementation workflows, not only raw star counts.

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?

This page now shows the strongest public repository signal available during the refresh. A high star count is useful context, but the editorial ranking also considers workflow fit, freshness, and whether AgentSkillsHub can produce practical implementation guidance.

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 are closest to current agent adoption pain points: teams need registry-backed MCP servers, inspectable browser execution, persistent memory, source-code context, and reusable workflow skills. Competitor scans and GSC data both support refreshing these clusters first.

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?

We plan to add historical trend charts. For now, bookmark this page and check weekly. You can also follow individual projects on GitHub and set up star notifications for projects you care about.

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

Guide

Best MCP Servers 2026

Quality-ranked directory (not trend-based).

Guide

Codex CLI Skills Directory

All Codex skills with install guides.

Guide

Neon Setup Guide (#1 This Week)

Get started with this week's top trending server.