Reject CLI tests that assert only an exit code

[jviotti]

Signed-off-by: Juan Cruz Viotti jv@jviotti.com

[augmentcode]

🤖 Augment PR Summary

Summary: This PR adds static validation to the CLI-test DSL so tests must observe produced results.

• Splits script parsing from execution through a reusable Statement representation.
• Parses WRITE ... UNTIL bodies before creating the sandbox or invoking the CLI.
• Adds a preflight data-flow check for RUN observations, generated artifacts, and checksum variables.
• Tracks comparisons, copies, and downstream input consumption to permit asserted output flows.
• Reports preflight diagnostics using the relevant DSL source line and stops before test execution.
• Retains the existing interpreter command dispatch, now operating on parsed statements.
• Extends decode and encode tests with assertions for previously unchecked captured output.
• Extends lint, install, test, and validate fixtures with explicit expected-output comparisons.
• Normalizes platform-dependent paths and benchmark timing values before newly added comparisons.
• Adds __pycache__/ to the repository ignore list.

_{🤖 Was this summary useful? React with 👍 or 👎}

[augmentcode]

Review completed. 2 suggestions posted.

Comment augment review to trigger a new review at any time.

[cubic-dev-ai]

All reported issues were addressed across 56 files

<sub>Reply with feedback, questions, or to request a fix.

Re-trigger cubic</sub>

[cubic-dev-ai]

All reported issues were addressed across 1 file (changes from recent commits).

<sub>Reply with feedback, questions, or to request a fix.

Re-trigger cubic</sub>