# Pilot product verification

Completed 7 September 2026.

- PDF: 20 pages generated with ReportLab, text extracted with pypdf, all pages rendered with Poppler and visually reviewed in five four-page sheets. No observed overlap, clipped text, stray blank pages or missing content. Cover background corrected before final render. Links and internal page references agree with the final pagination.
- Messages: 24 sequential unique IDs; all titles appear in the PDF. Editable text and PDF derive from the same content source.
- Spreadsheet: authored, recalculated, inspected, rendered and exported using the bundled artifact-tool library. Both sheet layouts reviewed. Nine behaviour checks covered the fictional example, changed inputs, zeros, inconsistent counts, fractions, missing prerequisites, middle and last rows, and restored blank state. See the included `tracker-checks.json` record.
- Export inspection: two sheet names, 26 formulas, waitlist dropdown validations and no exported error cells confirmed from the XLSX archive. Weekly input cells restored blank. Native Microsoft Excel interaction was not tested; rendering and calculation checks used artifact-tool.
- Customer ZIP: four expected files, CRC check passed, extracted bytes match originals. SHA-256 values are in `package-checks.json`. Seller drafts and internal source files are excluded.
- Sales copy: matches the file contents, clearly labels the US$29 price as proposed for the pilot, and contains no invented testimonials, scarcity or achieved outcomes. Seller identity, delivery setup, support and actual sale terms remain launch tasks.
- Research supports a concrete business problem and shows substantial free and paid competition. It does not establish willingness to buy this kit or its effectiveness. A practitioner review and buyer pilot remain necessary evidence before making stronger claims.

The source builders and machine check records are retained to make revisions reproducible. Rebuild and recheck affected artifacts after changes.