Specialized AI agents and tools are being deployed across teams, functions, and platforms — often without visibility and governance, creating the risk of shadow AI. Without a scalable way to find and track agents across multiple platforms, this sprawl makes it hard to understand not only what agents are operating in your organization but also how they reason, what they act on, and what outcomes they drive.
Salesforce’s enhancements to MuleSoft Agent Fabric bridge and simplify this fragmented landscape into a governed and cohesive agent network, giving you a single control plane for all your AI agents, tools, and metadata. This update centers around new Agent Scanners, giving you the ability to automatically detect and catalog AI agents across Salesforce Agentforce, Amazon Bedrock, Google Cloud’s Vertex AI, and other agent platforms. For other assets, like Model Context Protocol (MCP) servers and bespoke agents, additional capabilities make registration just as simple. By streamlining the integration of any agent or MCP server, regardless of vendor, Salesforce is delivering a trusted, open, and interoperable platform, empowering organizations to deploy the full potential of their Agentic Enterprise.
“The most successful organizations of the next decade will be those that harness the full diversity of the multicloud AI landscape. The expanded capabilities of MuleSoft Agent Fabric give you the freedom to innovate across any platform while maintaining the unified visibility and control needed to scale.” — Andrew Comstock, SVP & GM, MuleSoft, Salesforce
How It Works: A Single Source of Truth for AI Agents and Tools
Starting with Agent Scanners, you go beyond manual discovery and get the visibility and technical context required to trust agents deployed across your enterprise:
· Continuous Multi-Ecosystem Agent Discovery: Instead of tracking down agents across platforms yourself, Agent Scanners do the legwork for you. You can simply link different environments, like Salesforce Agentforce, Amazon Bedrock, and Google Cloud’s Vertex AI, and the scanners start identifying running agents in minutes. They constantly patrol these ecosystems to spot new or updated agents, identify their endpoints, and understand what they’re designed to do.
· Deep Metadata Extraction: It’s not enough to know that an agent exists; you need to know what it’s able to access and action. Agent Scanners go beyond the surface of an agent and automatically extract its specific capabilities (e.g., querying a database or processing a refund), the large language models (LLMs) powering it, and the data it has permission to access where available. They then normalize and map the metadata to Google Cloud’s standard agent-to-agent (A2A) protocol card specifications.
· Seamless Cataloging: The days of stale spreadsheets are over. Everything that Agent Scanners find is continuously synced to MuleSoft Agent Registry, a central catalog where agents, MCP servers, and AI tools can be registered and made discoverable by developers or other agents. It’s an always-on approach to cataloging agents that ensures your security teams are always looking at real-time data, not a snapshot from three months ago.
It’s not just AI agents on major platforms that can be automatically discovered. To help ensure no AI asset is left behind, MuleSoft Agent Fabric also includes flexible registration for homegrown agents and MCP servers via URL, plus a curated list of public MCP servers from the Official MCP Registry.
See What’s Next in Tech With the Fast Forward Newsletter
Tweets From @varindiamag
Nothing to see here - yet
When they Tweet, their Tweets will show up here.
