feat(restore): supercharge — elapsed_ms, exit_code, INTERNAL_ERROR fix, full docstrings
- Add elapsed_ms (wall-clock ms) and exit_code to --json output so agents can budget time and branch on error type without parsing shell exit codes - Fix critical bug: restore_object() failure now exits 3 (INTERNAL_ERROR) instead of silently mapping to 1 (USER_ERROR) - Remove always-empty skipped field from _RestoreResult; replace with the two-field any_user_error / any_io_error distinction that drives exit code - Suppress per-file stderr prints in --json mode to keep stdout valid JSON - Add summary text line: 'Restored N file(s).' after non-JSON output - Add logger.debug / logger.error calls throughout run() - Expand _RestoreResult, _resolve_source_manifest, _resolve_file_path, and run() docstrings to cover all edge cases and the agent-facing JSON schema - Fix test_cmd_restore.py _sha() to emit sha256: prefix (was breaking 23/26) - Add 20-test supercharge suite covering: performance (elapsed_ms timing), data integrity (schema completeness, field types), INTERNAL_ERROR exit code 3, concurrent two-thread restore, source+staged combos, summary text output
0 comments
muse hub commit comment sha256:c7d2ed76a997734cf49aa0d76048a6507f0ee0ccf2b667eb3492018dc8f4797a --body "your comment"
No comments yet. Be the first to start the discussion.