Add run-level linkage across related decision evaluations

[MuttleyD]

Context

MIDAS currently evaluates decisions as individual governed events, with each evaluation producing its own envelope, audit chain, and outcome.

However, many real-world decisioning scenarios involve multiple related evaluations that belong to the same higher-level execution instance or business journey.

Problem

There is currently no first-class way to link related decision evaluations together as part of the same run.

This makes it harder to:

• understand how multiple decisions relate to the same end-to-end flow
• trace execution across a journey rather than only at single-decision level
• support future journey analytics and escalation tracking
• distinguish isolated evaluations from evaluations that are part of a broader execution path

Proposed solution

Introduce a first-class run-level linkage model for MIDAS.

A Run should represent a higher-level execution instance that can link multiple related decision evaluations over time.

Each evaluation envelope should be able to reference a run_id, allowing envelopes produced during the same journey or execution path to be grouped together.

Scope

• Define a canonical Run model
• Introduce a stable run_id
• Link evaluation envelopes to a run
• Ensure multiple envelopes can share the same run_id
• Expose run linkage through the API and/or envelope model
• Preserve backward compatibility for evaluations that do not participate in a run

Acceptance criteria

• A first-class Run model exists in the domain
• Evaluation envelopes can reference a run_id
• Multiple related evaluations can be linked to the same run_id
• Run linkage is deterministic and serialisable
• Existing single-evaluation flows continue to function without requiring run context
• Feature is documented