Specifications

Specifications is the project manual: every specification section on the project, grouped by CSI division.

Spec books are uploaded on the Documents page

The register has no upload of its own, and that is deliberate. Intake for every document type lives on one surface.

To get a spec book in:

1. Open Documents and click the Specifications card.
2. Click Upload to Specifications and upload the PDF. See Uploading.
3. Ingestion examines the PDF for CSI MasterFormat section boundaries. If it finds more than one, the book is split, and each section is written as its own document and processed separately. A book with one section or none is processed as a single document.
4. The sections appear here, already imported, once processing finishes.

The empty register says the same thing and takes you there: "The spec book is uploaded on the Documents page; sections appear here once it is processed." with a Go to Documents button.

The second way in is Autodesk Forma. A connected project mirrors the section list from Forma, and those rows arrive as metadata that you then import individually. See Integrations.

Reading the register

Rows are grouped by CSI division. Click a division header to expand it.

Above the list, three counts: Total sections, Divisions and Imported.

The toolbar has a search box over number, title or filename, plus filters for status and division, and a sort control. The result line states how many of the total are showing.

Sort by Number, Title or Last synced, each ascending or descending.

While any section is processing, the page refreshes itself every 30 seconds. Refresh in the header re-reads the register on demand.

A section whose division header is missing is grouped under its number plus (no division) rather than being dropped from the list.

Status

Each row carries one badge.

Badge	What it means
Imported	The section's PDF is in Struxen and has been processed.
Processing	Processing is running.
Not imported	A source PDF exists and has not been imported yet. Selectable.
No source PDF	There is no PDF to fetch. Not selectable, and not a failure.
Failed	Processing failed. The tooltip carries the error.
Import error	The import itself failed. Shown alongside the status badge, and only on rows that had a source to import.

No source PDF covers two different situations, and the tooltip says which:

• A metadata-only register row. Forma holds no source PDF for this section, so there is nothing to import. Spec PDFs enter Struxen through Documents.
• A row that predates the extended Forma spec sync. It can still gain a PDF reference on the next sync.

The status filter offers the same vocabulary: All statuses, Not imported, No source PDF, Processing, Imported, Processing failed, Import failed.

Importing sections from Forma

Importing copies each selected section's PDF from Forma into Struxen and queues it for AI processing. It is a paid, metered action.

1. Tick the sections you want. A division header ticks all of its rows. Sections already imported, already processing, or with no source PDF cannot be selected.
2. The action bar at the bottom shows the selection count and an estimated cost, with the page count and per-page rate in its tooltip.
3. Click Import N spec sections.
4. The confirmation states the estimate and what it covers: "AI vision processing (~$4.20 across 350 pages) will run on the selected sections. Credits are deducted from your project pool when each section completes."
5. Confirm. Each section's PDF is copied into Struxen and queued in the background, and progress appears on this page as sections complete.

The estimate is pages multiplied by a per-page rate of $0.012. It is an estimate, not a quoted price: what is actually deducted is the credits the processing consumes. Credits and how they are drawn are covered in Billing.

Import is disabled on the demo project, which is read only.

Sections that came from a spec book upload

A spec book uploaded through Documents is split by ingestion, and its sections land here already carrying their PDF. They show as Imported and there is nothing to buy: you have already processed them once, on upload.

The import action is for the Forma path, where the register knows a section exists before the PDF has been fetched.

Forma mirror

A project with the mirror on shows a banner reading Synced from Autodesk Forma, with the time of the last sync and a Sync now button. A project without it shows a Mirror from Forma link to project settings.

On the very first sync of a project, the register shows First sync running: "We're pulling spec section metadata from Forma. PDFs will be imported when you click Import."

Troubleshooting

The register is empty after uploading a spec book. The book has to finish processing before its sections exist. Check the document's status on the Document Center; see Processing.

A spec book uploaded but produced one document instead of many sections. Fewer than two CSI section boundaries were detected, so the book was processed as a single document. This is common with a scanned manual, because a scan has no text layer for the boundary detection to read. See Processing.

Every row says No source PDF. These are metadata-only register rows. There is no PDF in Forma behind them, so there is nothing to import. Upload the spec book on the Documents page instead.

A row will not tick. Rows that are already imported, currently processing, or have no source PDF are not selectable.

The imported sections have nothing in them when you ask STRUX. The section PDFs are probably scans. Text extraction caps at 100 pages for a specification, and reads nothing at all from a page with no text layer.

No specification sections match these filters. Click Clear filters in the inline message.