Skip to content

Our tool

SQuAT

Schedule Quality Assessment Tool

Run a Primavera P6 programme through 49 schedule-quality checks in the browser, against metrics you personalise — with the offending activities named behind every failed check, not just a number.

Open SQuAT
The SQuAT interface: presets down the left, the families of check across the top, and the drop area for a P6 XML export.

What it checks

schedule quality, measured against personalised metrics

A score, red/amber/green against every metric, and the count that produced it. Every failed check opens the list of activities behind it.

A scored schedule in SQuAT: 69 activities, an overall score of 50% passing with four passes, three warnings and seven fails, and a metric table showing each check's count against its threshold.
A sample programme, scored.

  • 49 schedule-quality checks

    Flags logic, float and constraint issues in one pass — and names the activities behind every one, so a failed check is a list to work through rather than a number to argue with.

    • Missing predecessors and successors
    • Leads, lags and FS share
    • Hard constraints
    • Negative and high float
    • Critical path test, CPLI and BEI
    • Calendar validation and baseline execution

  • AI narrative report

    The results written up in plain English, ready to go into a programme narrative. Ask follow-up questions about the schedule and get an answer grounded in the metrics rather than a guess.

    • A written report from the results
    • Questions about the schedule
    • Your own API key, kept in the browser
    • A log of exactly what was sent
    • Or point it at a model running locally

  • Metrics you personalise

    Every pass/fail threshold is adjustable and any metric can be switched off, so the assessment matches what this client asks for rather than a generic standard.

    • Adjustable thresholds
    • Metrics on or off
    • Saved presets
    • Saved schedules

Confidentiality

the schedule never leaves your machine

The file is read in the browser. Nothing is uploaded and no copy of the programme sits anywhere afterwards — which, for a programme under a confidentiality agreement, is the difference between running a check and having to ask permission first.

The report and questions are off unless you turn them on, run on your own API key, and keep a log of exactly what was sent. Point them at a model running locally and nothing leaves the machine at all.

Open SQuATAsk about access

Built and maintained by Cranborne Group Limited. Accounts are by sign-in.