Local declarations
MCP servers can be configured directly in desktop clients, IDEs, and coding tools outside a central inventory.
Available now · MCP discovery
Find supported MCP clients and server declarations.
A configured declaration is evidence that a client knows about a server. It does not prove the server is enabled, reachable, connected, authorized, or used.
Direct definition
What is MCP discovery?
MCP discovery identifies supported Model Context Protocol clients and declared servers from bounded configuration evidence on managed endpoints.
AxLoop knowledge base
The visibility problem
Evidence must come before control.
Evidence gaps
Configuration alone cannot prove a server ran or that a tool was invoked.
Relationship sprawl
Clients, servers, and services create a fast-changing layer that needs careful correlation.
How AxLoop helps today
Discover from the endpoint outward.
AxLoop identifies supported AI applications, agents, local models, runtimes, and MCP configuration evidence. Each finding keeps installed, configured, running, and independently verified states distinct.
- Identify supported software and configuration evidence
- Preserve device and user scope
- Monitor meaningful state changes
Questions teams ask
- Q1Which supported clients declare MCP servers?
- Q2Which server identities and transports are visible?
- Q3Is the evidence configured, running, or independently verified?
Common questions
Answers, briefly.
- What is the Model Context Protocol?
- MCP is an open protocol that lets AI clients such as assistants, IDEs, and agents connect to external tools and data sources through MCP servers. It standardizes how models call tools, which makes integration easy and adoption fast.
- Why does MCP need discovery?
- MCP servers are often configured locally in a JSON file on a developer's machine. They can expose file access, databases, SaaS APIs, or shell commands without passing through central review, so security teams need to know which exist and where.
- What does MCP discovery report?
- It identifies supported MCP clients, server declarations, transport type, safe executable identity, source, and device or user scope where available. It keeps configuration evidence separate from running or verified connection evidence.
- Configured versus connected: why does it matter?
- A server listed in a config file may never run. A running server may not be connected to anything sensitive. Separating these states keeps investigations accurate and avoids overstating risk.
Agent Interaction Observability