Closed beta · waitlist BrainScribe is pre-launch. AHPRA-registered specialists practising in Australia — join the first cohort
BrainScribeai
Overview›Changelog

What shipped, and when.

Closed-beta release notes for the BrainScribe engine. Newest first. Engine versioning tracks the scoring and interpretation library; the app shell tracks separately.

·
v0.6.0

Twenty-four reference scales live.

  • Closed-beta scale library expanded to twenty-four reference scales — coverage across mood and anxiety, cognition and informant, ADHD and developmental, trauma, substance use, function and capacity, child and family, and sleep, pain and wellbeing.
  • Five scales (PHQ-9, GAD-7, NPI-Q, ESS, IQCODE-SF) ship with full per-scale recommendation libraries and symptom-profile classifiers.
  • Nineteen further scales enter active validation — interpretation fragments authored, recommendation paragraphs pending clinical sign-off.
  • Item-response table now starts on its own page in PDF exports, with rating-scale legend and column header repeating across pages.
v0.4.1

Evidence-summary voice for all recommendations.

  • All 44 recommendation paragraphs across PHQ-9, GAD-7, NPI-Q, ESS, and IQCODE-SF rewritten from directive register to evidence-summary register with named literature citations.
  • Each paragraph now appends an automatic clinician-judgement qualifier — the platform summarises evidence; the clinician decides applicability for the patient in front of them.
  • NPI-Q apathy alert softened from directive to evidence-summary; safety-critical items (PHQ-9 item 9, NPI-Q psychosis or aggression at severity 2+) retain firm standards-compliance language.
  • Markdown italics rendered in the report editor and PDF export.
v0.3.7

Functional-impairment items + domain-specific alerts.

  • ESS driving-safety advisory now surfaces at the top of the report when item-level endorsement triggers it — not buried in recommendations.
  • IQCODE-SF "all cutpoints" view added — the report shows the cutpoint relative to community, primary-care, and memory-clinic populations side by side.
  • NPI-Q domain-specific alerts route to the correct clinical pathway — psychosis-aggression to safety review, apathy to executive-function review, irritability to caregiver-burden review.
  • PHQ-9 and GAD-7 question 10 (functional impairment) now drive a dedicated functional-impairment paragraph in the integrated draft.
  • New item_present fragment trigger mode added; NPI-Q multi-axis item_value support landed.
Older entries

The engine has been under quiet development since late 2025; pre-v0.3.7 release notes are internal-only. The publishable changelog begins with v0.3.7.

Read next
LibraryScalesGrouped by what they measure. Twenty-four reference scales live. OutputReportsHow the integrated draft is composed from the record. ReferenceFAQFourteen questions, grouped by what they're really about.