perceptdot/percept

Overview

An observability platform for AI agents that automatically discovers and recommends MCP servers. It includes built-in connectors for Google Analytics 4 (GA4) to monitor agent behavior.

Best for

Best for
Developers building AI agents who need basic observability with GA4 and MCP server discovery

Use cases

• Monitor agent performance and behavior in production
• Auto-discover available MCP servers for agent integration
• Track agent interactions with GA4 analytics

How to use

Install

npx -y @perceptdot/core

Tools exposed

• percept_discover
• percept_recommend
• percept_installed
• percept_roi_summary
• ga4_overview
• ga4_realtime
• ga4_top_pages
• ga4_events
• vercel_deployments
• vercel_latest_status
• vercel_projects
• github_prs
• github_pr_detail
• github_workflows
• github_issues
• sentry_issues
• sentry_issue_detail
• sentry_releases

Tested with

Claude Code, Cursor, Windsurf, Cline

Notes

An observability platform for AI agents that automatically discovers and recommends MCP servers. It includes built-in connectors for Google Analytics 4 (GA4) to monitor agent behavior.

0 stars on GitHub. Last updated 2026-06-07. Licensed MIT.

Use cases

• Monitor agent performance and behavior in production
• Auto-discover available MCP servers for agent integration
• Track agent interactions with GA4 analytics

Pros

• Automated discovery of MCP servers reduces manual configuration
• Built-in GA4 connector enables standard web analytics for agents
• Open-source codebase allows customization and self-hosting

Cons

• Very low community adoption (0 stars) suggests early-stage or limited trust
• Only documented connector is GA4, limiting integration options
• No clear documentation on agent types or supported observability metrics

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.