Policy authority
The user is the policy authority. The user owns the Action Plane — its data, systems, and infrastructure — and defines what agents are permitted to do within it. The security policy:- MUST be defined by the user (or by a policy management engine acting on the user’s behalf).
- MUST be bound to the agent’s verified identity as determined by the plane boundary enforcement point (see Plane boundary) — NOT by the agent’s self-declared
idfield. - MUST be enforced at the Action Plane boundary (the enforcement point), not by the agent itself.
- MAY support a default policy that applies when no agent-specific policy is defined.
The agent does not participate in defining, negotiating, or interpreting the security policy. The agent is a subject of the policy, not an author of it. The agent cannot influence how it is identified to the Action Plane.
Policy scope
The security policy governs what operations an agent may perform. This includes but is not limited to:Server access
Which MCP servers the agent may connect to.
Scope grants
Which permission scopes are granted per server.
Resource constraints
Which files, directories, endpoints, or databases the agent may access.
EACCES when it violates it.
Standard errors
GPARS defines the following standard MCP-level error codes:| Code | Meaning | Returned by |
|---|---|---|
AUTHORIZATION_DENIED | The operation violates the user’s security policy. | Enforcement point |
SERVER_UNAVAILABLE | The target MCP server is not reachable. | Enforcement point |
data field:
error.code— a JSON-RPC error code (-32001for policy errors,-32002for availability errors).error.message— a human-readable description.error.data.gpars_code— the GPARS standard error code.