Golden Flows
The 5 normative integration scenarios that every MPLP-conformant runtime is expected to pass. These flows serve as the 'integration tests' for the protocol.
Any runtime claiming MPLP compliance MUST satisfy all Golden Flows.
Intent to Plan Transition
The fundamental loop. An agent receives a high-level intent, decomposes it into a structured plan, executes tasks, and verifies the result.
Normative Scope: Context · Plan · Trace
Governed Execution
Ensures every action is checked against policy before execution. The AEL (Action Execution Loop) enforces permissions and constraints.
Normative Scope: Plan · Trace · Core
Multi-Agent Coordination Loop
How multiple agents with different roles collaborate on a shared objective using the Collab and Dialog modules.
Normative Scope: Collab · Role · Dialog
Drift Detection & Recovery
Detecting when an agent's execution diverges from the original plan and automatically triggering a recovery or replanning sequence.
Normative Scope: Core · Trace
Runtime Integration & External I/O
Connecting the protocol to external systems (IDEs, CI/CD, Tools) via the L4 Integration Layer, ensuring all side-effects are tracked.
Normative Scope: Extension · Network
Normative
Protocol Compliance Boundary
Golden Flows define the minimum executable scenarios required for MPLP compliance.
They are not examples, tutorials, or best practices.
Any runtime, framework, or tool claiming MPLP compatibility MUST successfully execute all Golden Flows without violation.