Open the nine worked examples and reference files →
1. Function tests
The nine published Lab cases cover merging, extracting, deleting, reordering, compression, common-metadata cleanup, PDF-to-JPG conversion, password protection and splitting. Automated checks compare fixture hashes, page labels and counts, image dimensions, selected metadata fields and password-opening behavior. They do not establish complete coverage of every tool or input.
2. Output review
The reference files are inspected with independent PDF-reading code as well as the libraries that create them. The raster references use a Node canvas with the shared presets; browser JPEG encoders can produce different bytes and appearance. Published previews let readers compare the synthetic scan results. A successful download alone is not evidence that a real document meets its requirements.
3. Privacy model
Pageivo is designed around browser-local document processing. The tool code reads selected files into browser memory and produces the result on the device. We keep analytics and advertising configuration separate from the file-processing path and do not intentionally send selected PDF or image contents to a Pageivo conversion server.
4. Limit tests
Browser tools depend on the user's device. The fixtures include a numbered ten-page PDF and synthetic image-based inputs. This is a bounded test set, not certification of large real documents, physical phones or every browser. Watermark and page-number placement, HEIC decoding and browser-specific download/print paths require their own visual checks. Available memory differs by device, so there is no universal maximum file-size claim.
5. Documentation review
Tool pages and guides are reviewed when behavior changes. We state material limitations, including when compression rasterizes pages, when interactive PDF features may not survive a rebuild, and when a privacy action such as metadata removal is not equivalent to complete document sanitization.