Catch the title block that drifted before the client does.

The set drifts one sheet at a time

A drawing set does not go out wrong all at once. It drifts. One sheet gets opened, one value gets retyped, and the field stops matching the other twenty-three.

• ISSUE_DATE is 2026-06-08 on most sheets and 08/06/26 on the four someone touched last Friday.
• PROJECT_NAME reads Riverside Mixed-Use on every sheet except three, where it is Riverside Mixed Use - no hyphen, and to the eye, identical.
• REVISION was bumped to C on the drawings but the sheet index still says B.
• A title block on sheet 14 of 24 still carries the previous job's PROJECT_NUMBER, pasted from a template and never changed.

None of these throws an error. The drawing plots. The set looks finished. The reviewer or the client finds it after issue, and now it is a reissue, a transmittal, and a conversation about quality control.

The manual check does not scale

One way to catch this today is to look. Open each layout, read the title block, and notice when one breaks the pattern. On three sheets that works. On thirty-six across four files it does not - the drift is a one-character difference on one sheet buried in a set, and attention runs out before the sheets do.

The half-automated way is ATTOUT or DATAEXTRACTION to a spreadsheet. You get the values in a grid, but it is a one-way read: you still eyeball the columns for the odd one out, then go back into AutoCAD and fix each block by hand with EATTEDIT. The export shows you something is off. It does not tell you which value is right, and it cannot apply the fix.

MORPHO_TITLE audits the whole set at once

Run MORPHO_TITLE. It scans every paper-space layout in the drawing - and, if you point it at them, other closed DWG files and an AutoCAD sheet set (.dst) - reads the title-block attributes, and lays the sheets out as a register: one row per sheet, one column per field. The same field lines up in one column whether the tag is PROJECT_NAME, PROGETTO, or PROYECTO, because the field is matched by meaning, not by tag spelling.

Then Check Issues runs. It is a fixed set of grounded checks, not a rule engine you configure:

• A project field whose value differs across sheets - including a difference that is only a trimmed space or a changed case.
• A field filled on some sheets and blank on others.
• Empty or duplicate SHEET_NUMBER values.
• A Total Sheets count that is lower than the number of sheets actually present, or simply wrong.
• A required role - CHECKED_BY, APPROVED_BY - left empty.
• An unfilled template placeholder still sitting in the block: [ Client ], [ Studio Name ], Drawing Title.
• ISSUE_DATE written in two different formats across the set.

Each finding names the sheets it affects. When there is a clear majority value and the outliers are real, writable, non-placeholder cells, MORPHO_TITLE proposes that majority and lets you apply it to only the outliers - the three typo'd sheets, not the twenty-one correct ones. It targets each block by its stable handle, not the layout name, so the fix lands on the right block even when two layouts share a name.

Nothing is written until you click Apply. The whole apply is one transaction and one Ctrl+Z. Cancel discards every staged edit. A skipped or failed edit is recorded with its reason in a local report, so you see exactly what changed and what did not.

The outliers, side by side

Why the audit catches what the eye misses

Riverside Mixed-Use and Riverside Mixed Use are one character apart and read identically on a printed sheet. The audit compares the actual strings - case and whitespace included - so the odd sheets surface instead of hiding. Dates are compared by format, padding-independent, because a reviewer rejects a set for inconsistent ISSUE_DATE formats whether or not either is wrong. This is string-level work the eye is bad at and the machine is good at.

The field-driven case is where it stays out of your way. If a SHEET_NUMBER is driven by Sheet Set Manager, MORPHO_TITLE detects the field, locks the cell, and tells you why - so you fix it at the SSM source instead of typing plain text over a field the next regen would wipe out, severing your numbering automation.

Who it is for

It is for the person who owns the set before it ships - the CAD manager, the BIM coordinator, the drafter running pre-issue QA. If your title blocks are plain attributes and you have ever shipped a set with a drifted date or a stale project number, this is built for that exact moment, with zero setup: no .dst, no field pipeline, no spreadsheet template to prepare first.

It is the wrong tool if you want it to redraw or restyle the title block - it edits attribute values and reports inconsistency, it does not author geometry or invent data. It is also not a generic rule engine: the checks are a fixed, curated set chosen to avoid false positives, not a place to write your own validation logic.

FAQ

Does it change my drawings without asking? No. Every fix is staged and shown first. Nothing is written until you click Apply, the whole apply is a single Ctrl+Z, and Cancel discards everything staged.

Will it overwrite a Sheet Set Manager sheet number or an automatic date? No. Field-driven attributes, constant attributes, and xref title blocks are detected and locked - read-only in the scan, the grid, and the apply. If it cannot read a cell with certainty, it locks it rather than risk corrupting it. For a field-driven value the correct path is editing at the SSM source, and it tells you so.

Does it work across multiple drawing files? It audits the open drawing's sheets, and you can pull in other closed DWG files and an AutoCAD .dst sheet set as read-only rows in the same register. Write-back to closed files is a separate, guarded capability, not the headline flow.

Does it understand title blocks in other languages? Yes, for field matching. PROGETTO, PROYECTO, and PROJEKT all map to one Project Name column across EN/IT/ES/DE/FR/PT/TR variants. The match is a heuristic best guess; if it maps a column wrong, you reclassify it and the correction is remembered for that title block.

Can I write my own consistency rules? No. The checks are a fixed, grounded catalog - field drift, empty and duplicate sheet numbers, Total Sheets, required people, unfilled placeholders, date-format drift, and tracked raw columns. This is deliberate: a fixed set avoids the false positives a free-form rule engine produces.

Does it use AI or eat my quota? No. Detection and field mapping are deterministic heuristics - no AI call, no quota consumed. It runs on every plan tier, the same as the visual-style and standards-check commands.