v1.0: make transformation policies more explicit and inspectable #38
Description
While thinking through the v1.0 shape of axiom-core, one thing that keeps coming up is how visible and inspectable transformation policies are during development.
Right now, it’s clear what the SDK does conceptually, but when you’re tuning abstraction for a specific domain, it’s not always obvious why a certain piece of context was kept or removed unless you step through the code.
It might be worth adding a lightweight way to:
- surface which rules or policy decisions were applied during a transformation
- optionally emit a structured “decision trace” alongside the transformed output (off by default)
- make it easier to reason about tradeoffs between privacy strictness and reasoning quality
This feels especially useful for v1.0 users who need to explain behavior to security or compliance folks, not just engineers.
Not proposing a concrete API yet - mostly flagging this as a usability / transparency gap we may want to address before calling v1.0 “stable”.