Agent context should show what not to infer

Agent context should show what not to infer.

That may be as important as what to extract.

A public note has edges.

It does not carry every source.

It does not grant access to what stayed offstage.

It does not turn a question into a settled belief.

If an agent cannot see those limits, it will fill the gaps.

The context should make the negative space visible.

Here is what this note says.

Here is what it does not say.