You may have a cover, main report, and appendix as separate PDFs and need to combine them, add continuous page numbers after merging, but leave the cover unnumbered.
A submission-ready document may also need unnecessary pages removed, page order corrected, or a CONFIDENTIAL or DRAFT watermark added.
You can do each job in a separate PDF tool, but a longer workflow means repeatedly saving the document, reopening it in another tool, applying the next change, and saving again.
Browser Kitty's PDF Pipeline Builder connects these operations as nodes so the entire process becomes one pipeline: Merge → page cleanup → split cover from body → number only the body → merge again → add watermark → save. PDF processing runs in the browser, and the selected PDFs are not uploaded to Browser Kitty for conversion.
The PDF we will build
For this example, use three files: cover.pdf for the cover, report.pdf for the main body, and appendix.pdf for supporting material. They will be combined in the order Cover → Report → Appendix.
The cover will stay unnumbered. The first page of the report will be numbered 1, the next page 2, and the numbering will continue through the appendix.
Finally, a CONFIDENTIAL watermark will be applied to every page of the completed PDF. The important point is that merging and numbering are not separate one-off actions: the final processing procedure itself becomes the pipeline.
- cover.pdf → cover
- report.pdf → main report
- appendix.pdf → appendix
- Cover → no page number
- First report page → 1, second report page → 2, appendix → continuing numbers
At Split, the cover and remaining pages branch; only the body goes through Page Numbers before the streams are merged again.
1. Add three PDF Input nodes
Start by placing three PDF Input nodes and loading one PDF into each. Because this workflow combines three separate documents into one, it uses three inputs.
- Input 1 → cover.pdf
- Input 2 → report.pdf
- Input 3 → appendix.pdf
2. Merge the three PDFs
Add Merge Pages next. Merge Pages can concatenate two to six input streams in a defined order. Connect cover.pdf to A, report.pdf to B, and appendix.pdf to C so the resulting page stream is Cover → Report → Appendix.
When several PDFs should behave as one document with continuous numbering, merge them first and apply page numbers afterward. If each PDF is numbered before the merge, numbering begins independently inside each file.
Because the processing order is visible on the pipeline, it is easy to confirm whether numbering happens before or after the merge.
3. Remove unnecessary pages after merging
If report.pdf ends with an unwanted blank page, add Delete Pages after Merge.
Page expressions are evaluated against the page stream that reaches that node. When Delete Pages is placed after Merge, its page numbers refer to the already-combined document.
For example, use 8 to remove the eighth page of the merged stream, 8,12 for two separate pages, or 8-10 for a continuous range.
4. Reorder pages before adding page numbers
If page order needs to change, use Select / Reorder before Page Numbers. Nodes execute from upstream to downstream, so Reorder → Page Numbers assigns numbers to the reordered sequence.
Page Numbers → Reorder instead moves pages that have already been numbered. It is generally easier to finish page deletion and reordering first, then apply numbering to the final structure.
5. Split the cover from the body
Because the cover should not receive a page number, use Split Pages. Assume the cleaned merged document is page 1 Cover, pages 2 onward Body and Appendix.
Set Split Pages to 1. The first page goes to Selected, while page 2 and later go to Rest.
This does not save two separate PDF files. The processing path branches only inside the pipeline and can be joined again after the required operations.
6. Send only the body through Page Numbers
Connect only the Rest output of Split Pages to Page Numbers and set the starting number to 1. The original second page—the first page of the report—now receives number 1, followed by 2, 3, 4, and so on. The cover never passes through Page Numbers, so it remains unnumbered.
Number formats include 1, 1 / 10, Page 1, and Page 1 / 10. You can place the number at the top-left, top-center, top-right, bottom-left, bottom-center, or bottom-right, and adjust font size and margin.
7. Merge the cover and numbered body again
At this point the workflow has two streams: the cover, and the numbered body plus appendix. Add another Merge Pages node.
Connect the cover to A and the numbered body to B. The result becomes one PDF again: unnumbered cover, report page 1 numbered 1, report page 2 numbered 2, and continuous numbering through the appendix.
This ability to route only selected pages through a different process and then return them to the original flow is one of the useful parts of the pipeline model.
8. Add CONFIDENTIAL to every page
After the branches are merged again, add Watermark and set the text to CONFIDENTIAL. Watermark lets you adjust the text, opacity, angle, position, font size, and margin. Custom Watermark text currently supports 1–60 ASCII letters, numbers, and common punctuation characters.
This example uses Split → Page Numbers → Merge → Watermark. Putting Watermark after the second merge applies the same mark to the entire finished PDF, including the cover.
If the cover should stay clean, route only the body branch through Watermark instead. The same node affects different pages depending on where it is placed in the pipeline.
9. Preview intermediate results
As the workflow grows, you may want to check whether the merge order is correct, Split selected the intended page, or numbering really begins at 1.
Selecting a node previews the PDF state at that point. Merge can be checked for Cover → Report → Appendix order, Split can switch between Selected and Rest, and Page Numbers can confirm that only the body is numbered.
This lets you verify each stage while building instead of discovering a mistake only after the entire pipeline runs. Preview PDFs are also generated locally in the browser.
10. Connect PDF Output and run the pipeline
Connect the final stream to PDF Output. Press Run Pipeline to execute the nodes in connection order. Check the final result in Preview, then save the PDF when it is correct.
Node placement and connection order define the PDF processing order.
Save the workflow as a Recipe when you repeat it
If the same workflow is used repeatedly, save the completed pipeline as a Recipe. A monthly process might combine a cover, monthly report, and appendix, number everything except the cover, add CONFIDENTIAL, and save the result.
You do not need to rebuild the nodes every month. Open the Recipe and choose the current cover for Input 1, current report for Input 2, and current appendix for Input 3.
A Recipe stores the nodes, their connections, and node settings. It does not contain the source PDFs or generated PDF, which makes it suitable for applying the same procedure to new documents.
Run a Recipe without replacing the current Canvas
Saved Recipes can either be applied to the Canvas or run with Use this Recipe. The latter lets you choose PDF inputs and generate an output Preview without replacing the pipeline currently being edited.
This makes a frequently used PDF workflow behave almost like a dedicated tool. For recurring documents, the pipeline does not need to be rebuilt or edited every time.
Use Pipeline JSON to move the workflow to another computer
Recipes are stored in the current browser. To move a workflow to another computer or keep a portable backup, export it as Pipeline JSON.
Pipeline JSON stores nodes, connections, settings, node positions, viewport, and output filename. It does not include PDF file data, so each PDF Input must be selected again after opening the JSON in another environment.
How this differs from using one PDF tool at a time
Dedicated tools exist for merging PDFs, deleting pages, adding page numbers, or adding watermarks. If you only need one simple operation, a dedicated tool may be sufficient.
PDF Pipeline Builder is useful when several operations need to be combined, such as Merge → clean up pages → process only part of the document → merge again → finish the whole document. The workflow can then be saved as a Recipe when it needs to be repeated.
The main difference is that not only individual operations, but also their execution order, can be reused.
PDF processing stays in the browser
This example uses several PDFs: a cover, main report, and appendix. PDF Pipeline Builder processes selected files, intermediate Previews, Recipe execution, and generated PDFs in the browser.
The selected PDFs are not uploaded to an external server for processing. A standalone single-HTML build can also be saved and used offline.
Things to check when modifying PDFs
PDF Pipeline Builder rebuilds the output document from pages. Document-level information such as bookmarks, attachments, forms, digital signatures, and outlines may therefore not be preserved.
Modifying a digitally signed PDF can also invalidate the signature. Keep the original when a document is important and verify the generated PDF before using it.
Combine multiple PDF operations into one procedure
The workflow in this guide is: load multiple PDFs → merge them → clean up unwanted pages and order → split cover from body → add continuous numbers only to the body → merge cover and body again → add a watermark to the whole document → save as PDF.
Instead of reopening the document in several separate tools, the complete procedure can remain as one pipeline. Intermediate nodes can be checked as you build, and the finished workflow can be saved as a Recipe and reused with other PDFs.
PDF Pipeline Builder is less about one isolated merge or numbering action and more about building a repeatable procedure from several PDF operations.
Workflow at a glance
- Add three PDF Input nodes and load the cover, report, and appendix.
- Use Merge Pages to combine them in Cover → Report → Appendix order.
- If needed, clean up unwanted pages and ordering with Delete Pages and Select / Reorder.
- Use Split Pages to branch page 1, the cover, from page 2 onward.
- Send only the body branch through Page Numbers and start numbering at 1.
- Merge the cover and numbered body again with Merge Pages.
- Apply a Watermark such as CONFIDENTIAL to the completed document.
- Select intermediate nodes and use Preview to verify merge order and page numbering.
- Connect PDF Output, run the pipeline, and save the generated PDF.
- Save the workflow as a Recipe for repeated use or Pipeline JSON for portable backup.
PDF Pipeline Builder
Build PDF processing as a node pipeline, inspect intermediate results, and reuse saved Recipes with other PDFs.
Tips and limitations
- For continuous numbering, finalize merging, deletion, and reordering before placing Page Numbers.
- When the cover or another subset needs different processing, use Split Pages and Merge Pages to branch and reconnect the stream.
- For long pipelines, preview intermediate nodes so mistakes can be found at the stage where they occur.
Frequently asked questions
Can I merge several PDFs and then add continuous numbering from 1?
Yes. Merge the PDF Input streams first, then place Page Numbers afterward so numbering is based on the combined page sequence.
Can I leave only the cover unnumbered?
Yes. Split the cover into Selected and the remaining pages into Rest, send only Rest through Page Numbers, then merge the branches again.
How many inputs can Merge Pages combine?
Merge Pages currently combines two to six streams in input order. For a larger workflow, multiple Merge Pages nodes can be chained in stages.
Can I use the same pipeline with different PDFs?
Yes. Save it as a Recipe for reuse in the same browser, or as Pipeline JSON for backup and transfer. Neither stores the source PDF data, so input files must be selected again.
Are the selected PDFs uploaded to a server?
No. The hosted version first loads the application HTML, but selected PDFs, intermediate previews, Recipe execution, and output generation are processed in the browser and are not sent out for conversion.
Will bookmarks and digital signatures always remain intact?
Not necessarily. Rebuilding pages may drop document-level information such as bookmarks, attachments, forms, digital signatures, and outlines. Editing a digitally signed PDF can also invalidate the signature.