Find it. Trust it. Run it.
The trusted marketplace for AI agents, MCP servers, and skills. Security-scanned with a 9-point check. Verified publishers. One-click install.
Staff Picks
Hand-picked for quality and trust
Filesystem MCP Server
Secure file system access with sandboxing
GitHub MCP Server
Full GitHub API access for AI agents
Brave Search MCP Server
Web and local search via Brave API
Memory MCP Server
Persistent memory with knowledge graphs
Sentry MCP Server
Monitor and debug errors with Sentry
CodeRabbit
AI-powered code review agent
Trending This Week
Most installed this week
Cursor Agent
AI-native code editor with agentic capabilities
CodeRabbit
AI-powered code review agent
Filesystem MCP Server
Secure file system access with sandboxing
CrewAI
Multi-agent orchestration framework
GitHub MCP Server
Full GitHub API access for AI agents
LangGraph
Stateful multi-actor agent framework
Browse by Category
Trust is not optional.
36.7% of public MCP servers have SSRF vulnerabilities. 13% of community skills contain critical security issues. AgentXchange scans every listing before you install it.
9-Point Security Scan
SSRF, prompt injection, data exfiltration, dangerous commands, secrets, obfuscation, external fetches, credential access, privilege escalation.
Trust Score (0–100)
Transparent, multi-dimensional quality score combining security, quality, maintenance, and community signals.
One-Click Install
Install in Claude Desktop, Cursor, VS Code, or any MCP-compatible client with a single click.