tools/call
tools/call is the MCP protocol method that an agent client sends to an MCP server to invoke a specific tool by name, passing input parameters defined in the tool's schema.
Why this matters for AI agents
tools/call is where agent intent becomes action. It is the moment an agent reads data, writes a file, calls an API, or executes code. Enforcement here is enforcement at the point of consequence.
How AgntID relates
AgntID's primary enforcement point is the tools/call boundary—every request is evaluated against policy before being forwarded or denied.
The static IAM gap
MCP's tools/call method invokes whatever tool the authenticated client requests. There is no native per-call authorization layer.
Related phrases
Related terms
Frequently asked questions
What is tools/call in MCP?
tools/call is the MCP protocol method that an agent client sends to an MCP server to invoke a specific tool by name, passing input parameters defined in the tool's schema.
Why is tools/call the most important enforcement point?
tools/call is where agent intent becomes action—the moment an agent reads data, writes a file, calls an API, or executes code. Enforcement here is enforcement at the point of consequence.
How does AgntID enforce authorization at tools/call?
AgntID's primary enforcement point is the tools/call boundary. Every tools/call request is evaluated against policy before being forwarded or denied. This is where per-tool authorization happens.
Secure every agent tool call at execution time.
AgntID gives infrastructure teams scoped, ephemeral access control for AI agents without replacing IAM, MCP servers, tools, or agent frameworks.