Q: Something I’ve been trying to embrace with my tooling is to try to harness the creative problem solving of agents in a safe way. Do you think something like that is orthogonal to Onto or do you think there is a place in Onto where creative problem solving is still allowed? I don’t have a feel yet for how rigid the business rules and process modeling inside Onto would be.

Not orthogonal — Onto is a safety harness for agent creativity. The principle: creativity anywhere, authority nowhere but the model

Where agents get to be creative:

• Authoring (the big playground): coaching questions, decomposing stories, suggesting rules and fields. Anything goes, because every proposal must parse, resolve, and survive a human promote. Invalid output can't land. This is the main place for unbounded Agent creative thought.
• Runtime: Agents act as actors, choosing moves inside the legal set the rules define. A bad move is refused, not committed.
• Bespoke UI: Agents generate real running interfaces (seat maps, boards) against a contract that fenced all logic out.
• Agentic Work: A real example is an investment banking client we have - doing research and building decks within a set of rules

The one hard no: no improvising new logic at runtime. An unanticipated case becomes a refusal or a coaching moment

On rigidity: Onto isn't a workflow engine - no flowchart. Any action whose guard holds is available. The system is exactly as rigid as the rules you chose to write, no more.

<aside> 💡

Agents propose freely, the model decides - creativity is unbounded, authority isn't.

</aside>