Docs / Concepts

Conformance

Conformance is the proof that an independent host handles protocol success and failure paths consistently.

Plain English

Conformance is how a host proves it behaves like CHP says it should.

Why it exists

Public protocol surfaces need repeatable checks for malformed input, unavailable capabilities, version mismatch, policy denial, and evidence.

Formal definition

Conformance is a suite of protocol checks that verifies manifest validation, lifecycle enforcement, authorization behavior, structured outcomes, and evidence semantics.

Concrete example

Ground the concept before the schema.

A ServiceOpsHost conformance run proves that unknown capabilities and approval_required invocations return structured outcomes.

schedule_technicianVerified

Finds an available qualified technician and reserves a service window.

host: ServiceOpsHostpolicy: approval_requiredv1.0.0

Invocation trace

  1. 01 Actor

    Planning Agent

  2. 02 Capability

    schedule_technician

  3. 03 Host

    ServiceOpsHost

  4. 04 Policy

    manager_approval

  5. 05 Context

    job_context

  6. 06 Result

    Confirmed Appointment

conformance-case.json
json
{  "case": "approval_required_denial",  "expected": {    "outcome": "denied",    "success": false,    "denial": { "code": "approval_required" },    "event_type": "execution_denied"  }}

Relationships

Where this sits in the protocol.

Each concept should explain its neighbors so implementation teams can preserve the boundary across manifests, invocation, evidence, and tests.

Conformance depends on clear manifest and invocation semantics.

Infrastructure providers can run conformance against independent hosts.

Conformance findings feed trust, verification, and release decisions.

Visual model

  1. 01Test harness discovers a host manifest.
  2. 02Harness invokes valid and invalid cases.
  3. 03Harness records protocol outcomes and evidence shape.

Implementation notes

  • Test both success and failure paths.
  • Treat unknown hosts and unavailable capabilities as first-class cases.
  • Keep conformance output machine-readable.

Common mistakes

  • Testing only happy-path execution.
  • Assuming transport status is enough to prove protocol behavior.
  • Skipping authorization and lifecycle violations.

Related concepts

Keep reading through the boundary.

Host →Invocation →Policy →