What you provide
One unlocked PDF with at least one bookmark and up to 500 pages.
PDF and document workbench
The processor walks the document table of contents in stored hierarchy order.
Choose one unlocked PDF. The source remains unchanged and the result downloads separately.
Files are sent only when you run the tool; no server-side job history is created.
Product contract
The processor walks the document table of contents in stored hierarchy order.
One unlocked PDF with at least one bookmark and up to 500 pages.
One formula-safe CSV with outline level, title, destination page, and destination type.
Named destinations or unsupported actions may be classified broadly and require viewer inspection.
Workflow
Use PDF Bookmark Inventory on a working copy and review the separate result.
One unlocked PDF with at least one bookmark and up to 500 pages.
No hidden presets or inferred selections are applied; the complete supported structure is processed.
Open the download and verify its schema or document behavior. Named destinations or unsupported actions may be classified broadly and require viewer inspection.
Worked review example
A two-level outline produces one row for Overview and an indented child row for Details.
One unlocked PDF with at least one bookmark and up to 500 pages.
One formula-safe CSV with outline level, title, destination page, and destination type.
Named destinations or unsupported actions may be classified broadly and require viewer inspection.
Example output
A two-level outline produces one row for Overview and an indented child row for Details.
level,title,page,destination_type 1,Overview,1,internal 2,Details,2,internal
Implementation reference:PyMuPDF PDF API documentation
Common uses
Export the existing PDF outline as a reviewable CSV.
Review navigation before reorganizing pages.
Preserve an outline checklist before editing bookmarks.
Compare document editions for outline drift.
Important boundary
Named destinations or unsupported actions may be classified broadly and require viewer inspection.
Practical help
The processor walks the document table of contents in stored hierarchy order. It processes only the supported structures found in the selected PDF and returns a separate download. It does not overwrite the source, infer missing document intent, or certify legal, archival, accessibility, privacy, or security status. Review the artifact against the original before using it in another workflow.
One unlocked PDF with at least one bookmark and up to 500 pages. The request is limited to 10 MB, the document to 500 pages, discovered structures to 10,000 items, and generated output to 50 MB. Password-protected files must be unlocked first. Malformed, application-specific, or unsupported PDF objects may require a specialist desktop application and manual inspection.
Confirm the download opens and matches the stated output, then compare page count, visible content, navigation, form behavior, and any task-specific structure with the source. Named destinations or unsupported actions may be classified broadly and require viewer inspection. Keep the original until the result has been tested in the viewer, print process, records system, or downstream application where it will be used.
Document workflow
Export the existing PDF outline as a reviewable CSV. The processor walks the document table of contents in stored hierarchy order. One unlocked PDF with at least one bookmark and up to 500 pages.
One formula-safe CSV with outline level, title, destination page, and destination type. Named destinations or unsupported actions may be classified broadly and require viewer inspection. Completed downloads remain only in this browser tab for limited re-download.