Operational Reasoning.

The inversion.

An agent can carry workflow instructions in its system prompt and sequence steps in runtime code. Those instructions help it choose actions, but they should not be the only place the business rules are defined.

That can couple a business process to one agent implementation.

Models and runtimes change. When process rules are coupled to a particular agent, a migration can require reconstructing and retesting those rules.

Operational Reasoning externalizes the process contract. The schema defines entities, states, forms, allowed transitions, and audit structure. The model reasons over that schema and live entries, submits an activity, and the harness validates the action and records its outcome. Ordinary execution updates entries and history; changing the schema is a separate administrative action.

A compatible replacement model can read the same schema and submit the same activities. The contract and audit structure can persist, while tool compatibility, decision quality, and governance behavior still need evaluation.

The schema is the new prompt.

Prompts still express user intent and guide model behavior. Persistent schemas provide a separate, explicit contract for the workflow rules that must hold across different models and instructions.

A schema is different. A schema is a contract. It declares the entities, their states, the transitions between them, the activities that drive the transitions, the forms that capture the activities, and the audit shape that records them. The schema does not depend on which model is reading it.

In Inistate, the AI receives a user request and reads the schema and current records. It chooses an available activity and submits its form. The harness validates the submission against the schema and governance rules before committing the transition and recording history.

“The schema is the new prompt” is shorthand for moving process rules into a durable contract. It does not mean prompts disappear or that a schema performs the model’s reasoning.

In code, this is what it looks like.

An illustrative, abridged schema with one entity, four states, one activity, an AI confidence threshold, and the history-event shape. Only the submit activity is shown; this is not a complete importable module.

{
  "module": "leave_application",
  "entity": "LeaveApplication",
  "states": ["Draft", "Submitted", "Approved", "Confirmed"],
  "activities": [
    {
      "name": "submit",
      "actor": "any",
      "from": "Draft",
      "to": "Submitted",
      "form": {
        "fields": [
          { "name": "start_date", "type": "date", "required": true },
          { "name": "end_date",   "type": "date", "required": true },
          { "name": "reason",     "type": "string" }
        ]
      },
      "confidence_threshold": 0.85
    }
  ],
  "history_event": {
    "by":      { "type": "actor_ref" },
    "on":      { "type": "timestamp" },
    "changes": { "type": "field_diff[]" },
    "ai": {
      "type": "object",
      "fields": ["reasoning", "sources", "model", "model_version", "prompt_hash", "confidence"],
      "optional": true
    }
  }
}

This illustrative threshold uses a 0–1 scale: 0.85 means 85%, while the demo history displays confidence as percentages. Human and AI actions share the core history shape; AI actions add model and confidence metadata. An intention event records a proposed action that needs review, rather than a completed transition.

This is what externalizing operational intelligence into persistent schemas means in practice.

→ See the full schema

Why Operational Reasoning compounds.

A workflow schema can remain useful through multiple model generations. Reusing it preserves process definitions and audit conventions, while each replacement model still needs testing against the workflow.

Over time, an organization can accumulate schemas for leave applications, expense approvals, vendor onboarding, and incident response. Those definitions can evolve independently of the model while preserving audit continuity.

The compounding asset is the schema library. Models will continue to churn. The harness will continue to run.

This is why the right purchase decision is to buy by audit format and harness portability, not by the specific model your vendor happens to ship today.

→ Read the procurement argument