Gumloop Launches Gumstack for AI Agent Security and MCP Observability

Gumloop

Mar 20, 2026 · Updated Apr 25, 2026

Gumloop launched Gumstack, a security and observability platform for enterprise AI agent deployments. It tracks every tool call across all MCP servers, enforces RBAC and SSO, and provides full audit logs — available standalone or alongside Gumloop.

Gumloop, a no-code AI automation platform, launched Gumstack — its second product — as a security and observability layer for enterprise AI agent deployments. Gumstack gives teams a single dashboard to track every tool call across all Model Context Protocol (MCP) servers, see which users or agents initiated each call, and maintain full audit logs. It also includes an MCP inventory to spot unauthorized servers before they become a risk.

As AI agents multiply across enterprise environments, security teams lack visibility into what tools agents are calling. Gumstack addresses this by centralizing access control — RBAC and ABAC policies for MCP, SSO and SCIM via Okta or Entra ID, per-tool authorization, and credential management without local plaintext keys.

Teams running MCP servers across their org can centralize monitoring, enforce access policies per tool, and audit every agent interaction from one dashboard — with one-click deployment for internal servers.

View the full update on gumloop.com

Gumloop

@gumloopMar 18

Introducing our second product, Gumstack - a single pane of glass for AI agent data visibility and access control Track every tool call Manage every MCP server Full audit logs, RBAC, SSO Works standalone or with Gumloop https://t.co/voe5r1fsGg

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

Keep reading

Valeo Protocol Launches MCP Server for Autonomous AI Agent Payments

Valeo Protocol released an MCP server that integrates payments, credit, and budgeting directly into the AI agent workflow. This update allows agents to handle their own operating costs and financial tasks autonomously using standardized protocols.

CopilotKit🪁Apr 13

CopilotKit Launches MCP Server to Equip Coding Agents With Framework Expertise

CopilotKit released a dedicated Model Context Protocol server that provides AI coding agents with direct access to its APIs and best practices. By connecting tools like Cursor and Claude Code to this server, developers can eliminate hallucinations and ensure AI-generated code follows official implementation patterns.

OpenAI Launches Secure MCP Tunnels to Connect Agents to Private Data

OpenAIMay 27

OpenAI Launches Secure MCP Tunnels to Connect Agents to Private Data

OpenAI released Secure MCP Tunnel, a networking tool that lets AI agents access private on-premises databases and tools through outbound-only HTTPS connections. This removes the need for inbound firewall exceptions, allowing enterprises to safely ground agents in sensitive internal data.

LovableMay 8

Lovable Launches MCP Server to Build Full Stack Apps From Any Agent

Lovable released an official MCP server in research preview that allows external AI agents to programmatically create, edit, and deploy full-stack web applications. By exposing its app-building engine as a standardized toolset, Lovable enables developers to scaffold and ship production-ready software directly from their preferred terminal or coding environment.