Optimized `scripts/ci_validate_repo.py` to improve memory efficiency and execution speed:
1. Combined separate file-scanning passes into a single `validate_files` function.
2. Implemented early size check via `p.stat().st_size` to fail immediately on large files before reading.
3. Switched from `read_bytes()` to chunked binary reading (64KB) to ensure constant memory footprint regardless of file size.
4. Added robust error handling for file access.
These changes prevent the validator from loading large files into memory and reduce redundant I/O operations.
