| Folder selection shows fewer files than expected | The browser may skip protected or unsupported entries, or folder permissions were not granted. | Re-select an accessible directory, review hidden/system file expectations and compare the visible file count against the original. |
| A large folder scan is slow | Many candidates, large files, ZIP extraction or image similarity processing use browser CPU and memory. | Start with a smaller folder, try Standard for discovery, then Deep on the relevant candidates and avoid running multiple heavy tabs. |
| The scan is marked partial | Some contents were not read or hashed because of permissions, resource constraints or cancellation. | Treat the result as incomplete, narrow the selection and retry only the missing sources. |
| Two identical-looking photos were not grouped | Cropping, edits, rotation or threshold differences may exceed perceptual matching tolerances. | Review the pictures manually and retain originals; do not assume visual matching detects every transformed image. |
| Different photos are flagged as similar | Perceptual image hashes can be close for unrelated pictures with similar visual structure. | Preview both images before any cleanup and exclude false-positive groups. |
| Identical files with different names are not found in Quick mode | Quick compares names and can miss renamed byte-identical data. | Switch to Deep mode with content checksums. |
| Two files with the same name are grouped but differ | Name-only or name-and-size methods can produce candidate false positives. | Use Deep checksum verification and inspect contents. |
| RAR or 7Z archive cannot be read | The browser workflow primarily documents ZIP archive entry analysis. | Extract an authorized archive using a compatible extractor, then scan the extracted folder. |
| Password-protected ZIP cannot be scanned | The archive is encrypted or the selected extraction path does not support its encryption. | Unlock the archive with an authorized tool and scan readable files; avoid uploading confidential archives to unapproved services. |
| The browser freezes when reading huge photo libraries | Image decoding and grouping may exhaust available memory or CPU resources. | Split the library into smaller batches, close unused tabs and scan a representative subset first. |
| The tool reports space savings but disk free space did not change | In-page selections or reports do not physically remove stored files. | Confirm which files were actually deleted and inspect operating-system trash and cloud synchronization. |
| Deleted copies remain in their original folders | The in-page trash changes the review state, not necessarily the real filesystem. | Review the exported plan and perform an authorized disk cleanup separately only after backup and verification. |
| The wrong file was selected as keeper | Automated keep preferences follow rules rather than subjective document importance. | Choose manually, verify timestamps and quality and confirm required project paths. |
| I cannot find the undo option after running a script | External filesystem scripts may bypass the workbench recovery state. | Stop further cleanup and restore from a verified backup or available OS recovery facility. |
| Exported CSV report is empty | The scan may not have completed, no matching groups were found or a current filter excludes results. | Finish the scan, reset relevant filters and review on-screen groups before exporting again. |
| A deletion script includes a surprising path | Paths may be derived from selected folders, renamed entries or changed directory structures. | Do not run it; inspect each command, compare live file paths and regenerate the plan if necessary. |
| Cloud files do not appear in a local scan | Cloud-only placeholders may not be downloaded, or API access was not authorized. | Sync permitted files locally or use an approved API method after checking permissions and privacy terms. |
| A network drive cannot be selected | Browser file access and operating-system permissions may restrict mounted or protected shares. | Copy an authorized sample locally or use a compatible managed scanning workflow. |
| Some video or audio duplicates are missed | Re-encoded media can have different bytes; the advertised perceptual feature focuses on images. | Use specialist media-fingerprint comparison when content equivalence across encodings matters. |
| Similar versioned files were grouped incorrectly | Filename patterns such as final, copy and v2 do not prove revision lineage. | Compare actual document changes and ownership before retaining or removing any version. |
| ZIP scanning reports unexpected duplicate entries | An archive can legitimately contain copies across folders or repeated backup snapshots. | Inspect the ZIP internal paths and avoid modifying a backup solely to improve a duplicate score. |
| Storage estimate seems too high | Multiple copies, archive entries or hard-linked files can complicate naive savings estimates. | Compare estimated redundant bytes with actual filesystem space after any approved changes. |
| Unusual Unicode filenames look corrupted in exports | Character encoding or shell interpretation may differ between file systems and scripts. | Check UTF-8 handling and quoted paths, and test generated commands on safe copies. |
| The results change between two scans | File contents, metadata, source selection or scan mode may have changed. | Confirm the same sources, mode and hash algorithm before comparing reports. |