TestRunner

File: tckit/ports/test_runner.py Purpose: Run unit tests on the PLC runtime and return parsed results.

Method	Returns
run_tests(target_ams_id, *, plc_name=None)	Result
get_results(target_ams_id, *, plc_name=None)	TestResults

run_tests blocks until the suites finish or the bridge's timeout fires. get_results returns the parsed XML the run wrote — call it after run_tests has succeeded.

target_ams_id and plc_name mirror the IDE workflow: you pick both the target route and the test PLC project before running tests. Both are explicit on every call so an MCP session can't accidentally run on the wrong target through implicit state.

TestResults is a parsed suite/test tree with pass/fail and assertion detail (expected / actual / line) — not a console scrape.

Why this shape

The point of a test loop is feedback the model can act on without re-reading everything. A structured result tree lets it locate the one failing assertion and jump straight to that POU item via ProjectReader. Same tool-design principle as BuildRunner: parsed beats raw.