PDF 工作流构建器

It chains several PDF operations into one run. Instead of opening a tool, downloading, and feeding the result into the next tool for each step, you connect nodes on a canvas, such as PDF Input, Compress, Watermark, and Download, set each node's options, and click Run to process everything in sequence.

Yes. The PDF Input node accepts multiple files; a Merge node combines them into one document, while other nodes process each PDF separately. When more than one file comes out the end, the Download node bundles them into a ZIP automatically.

Sockets are typed: PDF (a single document), MultiPDF (several documents), and Image. A node that expects a single PDF won't accept a MultiPDF output, so put a Merge node in between, and image outputs only feed nodes that take images. Connections always run from the bottom of one node to the top of the next.

Once a PDF is encrypted, later nodes can't read it, so the engine rejects a workflow where anything other than the output sits after Encrypt. Move Encrypt to just before Download. Password-protected inputs are the opposite case: the run pauses and asks for the password.

The run stops at that node, which turns red with an error message, and nothing downstream is produced. Fix the setting or the input and run again; it's easiest to build a two-node version first and add steps once that works.

Yes. Save stores it as a named template in your browser's local storage, and Load brings it back in a later session on the same browser. Export writes the layout to a workflow.json file you can back up or send to a colleague, and Import restores nodes, positions, connections, and settings from that file.

No. Templates and exported JSON files hold only the node types, their settings, positions, and connections; the PDFs are chosen fresh each time you run. Settings are saved as typed, so avoid exporting a workflow with a password entered in an Encrypt node if you plan to share it.

No. Every node runs in your browser with the same engines as the individual tools, and the only network activity is fetching a processing module such as Ghostscript the first time a node needs one. Your PDFs never leave the machine, even in a long multi-step run.