rc_ok treats mixed sanitizer failures as tolerated leaks

[Nitjsefnie]

Describe the bug
On current main, rc_ok() returns success for a nonzero command when its captured output contains both a LeakSanitizer report and a hard ASan or UBSan diagnostic. The command is counted as a tolerated leak, so the hard sanitizer failure is masked.

This is distinct from #945/#953, which covered tests/test_doc_examples.py. It was discovered during an independent review of unfiled sanitizer test-harness findings on 2026-08-18.

To reproduce
From the repository root, load the current helper and pass it mixed sanitizer output:

LEAKED=0
source <(sed -n '93,100p' tests/run_all_tests.sh)
output=$'ERROR: LeakSanitizer: detected memory leaks\nSUMMARY: AddressSanitizer: 16 byte(s) leaked in 1 allocation(s).\nruntime error: signed integer overflow'
rc_ok 1 "$output"
printf 'status=%s leaked=%s\n' "$?" "$LEAKED"

The command prints status=0 leaked=1.

Expected behavior
rc_ok() tolerates a nonzero exit only when the captured output represents a leak-only report. Any accompanying ASan or UBSan diagnostic keeps the command failed.

Actual behavior
The LeakSanitizer text is sufficient for rc_ok() to return success even when hard sanitizer diagnostics are present in the same output.

Environment

• OS: Linux; the classification is shell/grep control flow and is not architecture-specific
• GCC version: not applicable; the minimal reproduction does not compile the project
• EigenScript version: main at a916975b765581b6c531174ba28a67a05d5c277f