Verify example results for auto-testing

[trevorhardy]

Right now we have the Github actions running the examples and the workflow fails when any of the examples fails to complete. This is good and will catch changes in Python, PyHELICS or the HELICS API that are fundamentally breaking. What we probably need is some way of verifying the data actually coming out of the co-simulations is correct, though.

[Gaurav890]

I opened #137 with a focused first slice of result verification for the fundamental default example.

The proposed pattern adds a recorder as an observer federate, writes the exchanged values to a generated JSON artifact, and runs a separate pytest validator after the federation completes. This keeps validation logic out of the tutorial federates while checking both communication directions and model-level outcomes.

The initial checks cover all ten voltage/current publications, charging completion times, peak fleet load, and delivered energy. I kept the scope to one deterministic example so the pattern can be reviewed before applying it to additional examples.