Add malformed synthetic event-stream fixtures for parser regression

[bsmensah-ctrl]

Problem

The event-stream parser has good inline tests for malformed input, but contributors do not have reusable public fixtures that demonstrate the validation boundary.

Proposed change

Add a small invalid-fixture pack under fixtures/invalid/ and use it in parser regression tests.

Acceptance criteria

• Include at least three minimal JSONL fixtures covering:
  • malformed JSON;
  • a missing or empty required string field;
  • a negative or non-integer PID/tick field.
• Add a parametrized test that asserts the relevant InputError message for each fixture.
• Document the invalid fixture pack in docs/FIXTURES.md, including that these files are expected to fail validation.
• Keep every example fictional and synthetic; do not add live system data, acquisition code, or platform API access.
• uv run --frozen pytest -q and uv run --frozen ruff check . pass.

Useful starting points

• src/crossview_lab/simulator.py::load_event_stream
• tests/test_simulator.py
• docs/FIXTURES.md

[RubenSanosh]

Hi @bsmensah-ctrl, I'd like to work on this one! This will be one of my first open source contributions, so it might take me a little longer than usual to get up to speed, but I'll dig into the existing parser tests to understand the validation boundaries you mentioned before writing the fixtures. I'll comment here if I run into questions. Thanks for the clear write-up!

[bsmensah-ctrl]

Thanks @RubenSanosh — it’s yours. Start with one minimal fixture per validation boundary and keep the parametrized assertions focused on the public InputError message rather than parser internals. A draft PR once the first fixture and test pass is welcome; there’s no rush. Comment here if any acceptance criterion is unclear.