Exit cleanly after test failures

[liufengyun]

Summary

Failing Runner.runWithArgs runs append an exception traceback after the test summary. Exit with status 1 using Python's sys.exit and Ruby's Kernel.exit so the summary remains the final test report and normal runtime cleanup runs.

Add subprocess regression checks for both failure fixtures: status 1, the expected assertion details and summary, matching backend reports, and no runner traceback. Run these checks through the existing output-checking CI step, replacing the shell check that required the old tests failed exception message. Document the exit status in the README and record the fix under Unreleased in CHANGELOG.md.

Add a PR template with a test and changelog checklist and a compatibility section.

Closes #2.

Check list

• Add tests
• Update CHANGELOG.md

Validation passed: both runner conformance suites, output parity and clean-exit checks, and all three package builds. git diff --check passed.

Local validation used Jo 0.14.0, Python 3.10.12, and Ruby 3.0.2. CI targets Python 3.12 and Ruby 3.3.

Documentation changes were reviewed and checked with git diff --check; the runtime checks above passed for the implementation commit.

Compatibility

Public API signatures, package dependencies, and minimum runtime versions remain unchanged. Failed Runner.runWithArgs runs still report status 1, with the final abort exception and traceback removed from their output. Callers that need to handle failure themselves can use the Boolean returned by Runner.run.