Skip to content
View in Notion

Appraisals & Work Based Assessments (WBAs) ​

Feature Overview ​

WBAs are formal assessments used in medical training to evaluate clinical competency. PPP supports four WBA types (CBD, CEX, PBA, SSA, which is the UK DOPS equivalent) plus MCR; MSF and 360 reviews are Appraisal records, not WBA types (#12836). The "Yearly Appraisal" module (being renamed to simply "Appraisals") aggregates WBAs, reflections, and supervisor sign-offs into an annual portfolio.

The appraisal system provides a single digital repository for career progression, accreditation, peer-to-peer learning, and performance tracking across training levels.† It encompasses MSF (collated by one consultant from multiple sources), MCR (multiple consultants reporting on non-consultants), 360-degree feedback (medical and non-medical staff), and a Yearly Appraisal.†

Open Questions ​

#QuestionSourceAnswer
Q20Should colleagues' MCR response time count working days only (current: 7 working days ≈ 9–11 calendar days) or calendar days (7 days, exactly one week)? Narrowed 2026-10-03 to part (a). Matt said in 2026-03 that 48 hours was too tight when a report is started on a Friday, and on 2026-06-03 answered "yes" to extending it. Rob set 7 business days, settings-driven, as acting PO on 2026-06-30; Matt asked for the outcome on 2026-06-27 and 2026-06-29 and has not replied since. (The "Matt's settled decision" label on #9059 was Rob's account, not Matt's words.) Parts (b), what happens when the window closes, and (c), whether 360 has a window, are closed (see Closed Q20 (b, c)).†transcript-2026-03-11 (~01:02:01); issue-8170 (2026-06-03); issue-9059 (Rob 2026-06-30)OPEN — awaiting Matt. The 7-working-day default stands until he answers. Engineering lean: working days, from his own Friday-to-Monday example.
Q21Reflections model — does the Pendleton feedback model apply to the Reflections form? The prior AC "Reflections follow Pendleton feedback model for structured retrospective practice" was inferred from transcript-2026-05-19 §PendletonModel (~00:43:00) — "platform supports the Pendleton model of feedback to create safe, implementable feedback loops". Investigation (#9827) found this transcript quote describes the Feedback tool context, not the Reflections tool. Engineering default: Pendleton applies to the Feedback tool only — not Reflections.#9827 / transcript-2026-05-19 §PendletonModelANSWERED (Matt D8178.8, 2026-07-13, #9653). Matt: "Free-text as per original spreadsheet, please agenda for call to run through current; This is a drift." † Reflections are free-text — Pendleton belongs to the Feedback tool only, not Reflections, confirming the engineering default. The source-of-truth "original spreadsheet" (GDrive Copy of PPP - Retrospective Evaluations.xlsx, Reflection Tool tab) defines the reflection as free-text fields Case Summary, Situation, Learning (each [Text Field]), plus Date, Hospital, and an optional Share-for-Review to a consultant — which is exactly the shipped ReflectionForm structure. No form change required; the current free-text form already matches. Matt flagged this thread as "a drift" and asked to agenda a call to run through the current build — that run-through sign-off remains open (issue keeps for-discussion), but it does not block confirming the free-text model. Do not add Pendleton-model fields to the Reflections form.
Q24Critical-condition level-4 rule: one level-4 assessment per condition, or a number? Matt asked on #9123 (2026-06-28) to "confirm all are just to be reached for level 4 for single assessment". The ISCP curricula say there must be documented level-4 CEX/CBD evidence for the critical conditions, with "no requirement for a certain number" (e.g. Cardiothoracic). Is the rule each critical condition reaches level 4 at least once? Non-surgical specialties are still TBA (Notion WBA brief).#9123 (Matt 2026-06-28); wba.competency.preloaded-horizontal-barsOPEN — awaiting Matt. Until answered, #12716 builds to the curriculum wording (per-condition level-4 evidence, no pooled count).
Q25Long-term competency trend: what should it look like? Matt's Task 9 brief asks for "Feature dashboard stage displays of increasing competency to signify long term trends". Should it be a per-type timeline of the global-summary level, a running count of level-4 assessments, or something else?#11194 (Matt 2026-07-29); wba.dashboard.competency-long-term-trendOPEN — awaiting Matt. Context, 2026-05-28 (Platform Demo transcript): Matt described a "longitudinal display of increasing competency over a level of training", running from level 0 (not enough evidence) and level 1 (assisting) up to level 4 (completely competent and independent). That fixes the scale but not the chart form.
Q26WBA cadence goals widget: position and display? The Notion WBA brief asks for a goals widget tied to the cadence meter ("how many to complete of which this 3 months") and marks its position as "for discussion": a horizontal runner, top or bottom?Notion WBA brief (last edited 2026-05-19); wba.widget.cadence-goalsOPEN — awaiting Matt.
Q27When a consultant signs off a WBA, or completes an MCR, MSF or 360 for a colleague, should it earn a fixed CPD amount per form under supervision/teaching (e.g. 0.5 h), or count only as a logged involvement with no hours? And if it earns CPD, in which category? Matt asks for "CPD hour triggers where appropriate for Consultant user" (#11194), "automated CPD scoring" (Notion WBA brief) and the "accounting and logging of any triggers CPD hours/points" for appraisals (#11210), but no source gives the amount or the category. Q29 (MCR/MSF/360 completers) was folded in here on 2026-10-03. (Logbook sign-off is different: consultant CPD there is already built but drifts from Matt's #9200 rules — see logbook.consultant-review.cpd-credit-on-sign-off, #12725.)#11194 (Matt 2026-07-29); #11210 (Matt 2026-07-29); Notion WBA brief; wba.cpd.consultant-sign-off-credit; appraisals.lifecycle.completer-cpd-creditOPEN — awaiting Matt. Build tracked in #12712. Context, 2026-05-28 (Platform Demo transcript): Matt said senior users' WBA involvements are "supervisory and educational in nature" and "would then score against their CPD home". He gave no hours and named no CPD category.
Q28Yearly-appraisal PDP: what shape should the long-term plan and next-year steps take? Matt's 28 Jun brief splits the PDP into (a) next year's CPD plan, (b) a long-term plan and (c) steps for next year tied to it. On the same day he left a Notion comment asking Rob to "include reference to the L2P appraisal system previous provide which will assist in development of b and c". Rob replied (2026-07-01) that he still needed to find the L2P material. It is not in the requirements corpus and has not been located. Without it, the fields for (b) and (c) can't be designed from Matt's sources.Notion roadmap/appraisal/runner brief, inline comment 2026-06-28; transcript 2026-07-01 (~00:54:06); appraisals.yearly.pdp-three-partsOPEN — needs the L2P reference from Matt (or its Drive location). Until then pdp-three-parts is specified at the level Matt described: three parts, with step types publications, presentations, exams, conferences and courses. No field-level design yet.
Closed Questions (24) — click to expand
#QuestionSourceAnswer
Q1MSF / MCR / 360 — are they user-initiated or department-scheduled?Appraisal briefAnswered 2026-05-03 (Matt): User-initiated. Department scheduling deferred.
Q2Yearly Appraisal lifecycle — draft → review-with-appraiser → signed-off → locked; or simpler?Yearly Appraisal briefAnswered 2026-05-03 (Matt): 3 phases — Draft → Review with appraiser → Signed off. Signed off = locked (no separate 4th state). See Yearly Appraisal Lifecycle below.
Q3Critical-condition WBAs vary per speciality — where is the master list maintained?WBA briefAnswered 2026-05-03 (Matt): PPP Admin account (ops-editable). Currently orthopaedics only; needs expanding for all surgical specialities using same source docs as logbook extraction. Refined 2026-07-14 (Matt, transcript): in the admin console, split the specialty data into a CPD tab and a separate Training sign-off requirements tab. Per specialty, the second tab holds the principal procedures, the critical conditions, the number of level-4 PBAs per principal procedure, and the WBAs required for the critical conditions. This applies to non-consultants; consultants have the high-level specialist requirements.† The read-only review screen is specified in super-admin (super-admin.specialty-classification-review.*). Still incomplete as of 2026-09-29: Matt could not find the critical conditions in the admin console and said the extraction from the source PDFs was "still not there yet"†. The 29 Sep Weekly Tech meeting still had "extract specialty information from the WBA PDF" as an open action†. The data gap is #12713.
Q4WBA current web-forms — can Matt share the spreadsheets so engineering can data-model?WBA briefReframe (5 May). The original question was wrong-shaped. WBA forms are already built (forms/WBAForm/, forms/WBAConsultantForm/FormCBD.config.ts, FormCEX.config.ts, forms/MCRForm/, forms/MSFForm/; Dec 2025 screenshots wba_*.png/appraisals_*.png plus the Oct 2025 walkthrough video 39 - WBA (v1 Oct 2025).mp4). Form prompts are database-driven via Settings → Form Prompts (#2283); hardcoded values in code are fallbacks. The spreadsheet is a reference, not a requirements doc. Matt iterated heavily during the Moreton Blue UAT pilot and didn't always update the spreadsheet — so where existing forms diverge from the spreadsheet, we must NOT assume the spreadsheet is the source of truth. Each delta is a Matt-confirmation question, not an auto-correct. Example detected delta: FormCEX.config.ts rating items q6 = "Organization skills" / q7 = "Management and follow-up" vs spreadsheet's "Communication and listening" / "Organisation and time management" — could be Moreton Blue iteration, could be code drift; only Matt can say. Trail: 17 Apr brief flagged outstanding spreadsheet task; 28 Apr we asked; 3 May Matt asked "You don't have this from Moreton blue? or already in drive?"; Moreton-blue folder had only screenshots; 5 May Matt re-shared as native Google Sheets. Schema captured as § WBA Form Schema (audit reference). Drift-audit work tracked in #5720.
Q5DOPS (#5024) — build now or after Wave 6 split?Drift triage 2026-04Answered 2026-05-03 (Matt): DOPS = SSA — same thing, different name. Surface this equivalence in the UI to avoid confusion rather than building separately.
Q6Evidence-pack PDF — priority order for #5096 redirect, #5090 SVG bug, #5011 export?Drift triage 2026-04Resolved 5 May (code audit). Two of the three issues are stale; the answer is "no priority order — there's only one piece of real work". (1) #5090 already fixed — its SVG-blob bug describes a state of Reports/MultiFeature/utils/pdfExport.ts that was rewritten in #5229 (16 Apr) + #5271 (24 Apr). Current code uses @pdfme/generator to emit application/pdf Blobs and triggerDownload properly attaches+clicks an anchor. Closed. (2) #5011's "no PDF service exists" claim is stale — the codebase has certificateService, certificateGenerationService, and @pdfme/generator wired up; the evidence-pack PDF should be built by reusing the Multi-Feature pdfExport pattern, not from scratch. Description updated, kept open as the implementation issue. (3) #5096 is the symptom of #5011 not being built — there's no evidence-pack route anywhere in source, so React Router falls through to /mcr. Rescoped to a small frontend-only "coming soon" placeholder route, deliverable independently of #5011.
Q7MSF/360 rater anonymisation (#5008) — threshold (≥N raters) or fully anonymous?Drift triage 2026-04Answered 2026-05-03 (Matt): Not anonymous.
Q8360 minimum reviewer count (#5009) — statistical-validity minimum?Drift triage 2026-04Answered 2026-05-03 (Matt): Minimum 7, plus Educational Supervisor (non-consultant) and Colleague Appraiser (consultant).
Q9Colleague Appraiser filter (#5063) — Consultant-only, or any consultant-role holder?Drift triage 2026-04Answered 2026-05-03 (Matt): Educational Supervisor (non-consultant profile) is synonymous with Colleague Appraiser (consultant profile).
Q17Appraisal form vs request appraisals buttons — are they distinct? Today's session (§8, 00:16:11) assumes they serve different purposes. Existing AC #5092 treats the dropdown as broken (always opens generic form). Does "request appraisals" open a type-selection flow, while "appraisal form" opens the current-appraisal wizard? Engineering needs to confirm in code before closing #5092. confirm-with-mattTranscript 2026-05-19 §8Answered 2026-05-19 (code audit): Yes — distinct. "Appraisal Form" (outlined button) opens the full annual appraisal wizard (AppraisalFormModule, lazy-loaded). "Request Appraisals" (contained button, + Request Appraisals) opens AppraisalTypeSelectionModal — an autocomplete picker for MCR / 360 / MSF — which then navigates to appraisals/{type}?add=1 to start a peer assessment request. Matt was asking what the buttons do; the answer is they serve entirely different journeys. #5092 can be re-evaluated: the "dropdown" issue may be about the type-selection modal, not the annual appraisal.
Q18Audit event tracking — QH-specific or all hospitals? Matt described audit events in the context of Queensland Health hospitals (§19, 00:45:53). Is the audit event → case participant notification scoped to QH-linked hospitals only, or should it work for any hospital in the system?Transcript 2026-05-19 §19Answered 2026-05-19 (Matt): Any hospital in the system — not QH-specific.
Q19"Three required level-4 assessments" — fixed count or specialty-specific? Matt said "track progress toward the three required level four assessments" (§16, 00:36:32). Is the number 3 fixed across all specialties, or does it vary by specialty like critical conditions?Transcript 2026-05-19 §16Answered 2026-05-19 (Matt): Varies by specialty. Source PDFs are in GDrive: Orthopaedic · General Surgery · Urology · ENT · Neurosurgery · Plastic · MaxFax · Cardiothoracic · Cardiothoracic (Congenital) · Paediatric · Vascular. Engineering must extract per-specialty counts before implementing the bar. Note: Vascular is in GDrive but not in local temp/.
Q10#5720 AUDIT — CEX ratings q6/q7 text mismatch. Code fallback had q6 = "Organization skills", q7 = "Management and follow-up", which near-duplicated q4 "Management and follow-up planning" and diverged from the CEX source sheet.#5720 / #6995Answered 2026-05-27 (Matt, #6995): Revert to source-sheet wording — q6 = "Communication and listening skills", q7 = "Organisation and time management". This also removes the near-duplicate with q4. Implemented in FormCEX.config.ts.
Q11#5720 AUDIT — MCR Part 1 has 9 capabilities in code; spec says "7+ capabilities". Items 8 (Education and training) and 9 (Research and scholarship) are GMC/UK extensions beyond AHPRA's 7 core domains.#5720 / #6998Answered 2026-05-26/27 (Matt, #6998): Keep all 9 for every doctor (AU/UK/NZ/Canada) by default — collecting more safeguards free trade of doctors across jurisdictions. Do NOT hard-remove 8 & 9; build a single off-switch so an AU enterprise can opt out. Implemented as MCR_SHOW_OPTIONAL_CAPABILITIES (default ON) gating items 8 & 9 in FormMCR.tsx.
Q12#5720 AUDIT — PBA/SSA Global Summary sub-levels not in code. Spec said PBA/SSA split each level into a/b sub-levels (9 total) but the global summary used the 5-level scale.#5720 / #7001Answered 2026-05-27 (Matt, #7001): Confirmed the 9-level picker (Level 0, 1a, 1b, 2a, 2b, 3a, 3b, 4a, 4b) for PBA and SSA only. Descriptor text matches the PBA1/SSA1 tabs verbatim. CEX/CBD stay 5-level; MCR stays 5-level. No code change required — FormSSAPBA.ReviewSection.tsx already consumes the 9-level a/b DOCTOR_RATINGS from constants/doctor-ratings, so the #5720 audit premise (that the global summary used the 5-level scale) was a false alarm; #7001 adds a regression test guarding the 9-level scale. Stored 5-level values on old records still display (the radio group does not clear unknown values).
Q13#5720 AUDIT — QMF/MSF end-section fields missing from config. The MSF form should close with (a) specific-examples free text, (b) probity & health concerns toggle, (c) environments-observed multi-select.#5720 / #6999Answered 2026-05-27 (Matt, #6999): "Lets build the three." Added all three closing fields to the live MSF form (FormMSF.config.ts + FormMSF.ReviewSection.tsx), wording from the Quarterly Multi Source Feedback tab. Environments options: Primary care / Specialist practice / A&E/admissions / Critical care / MDT / Outpatients / Theatre / Ward / Other. Also corrected the user-facing display name QMF → MSF (persisted wbaType: 'qmf' enum unchanged to avoid data migration).
Q14#5720 AUDIT — CEX "Focus of Encounter" options not in code. Spec lists options: History / Exam / Diagnosis / Management / Explanation / Consent. Code has conFocusOfEncounter: [] (empty array, no option list defined in config). Are these options currently seeded in the DB, or do they live elsewhere in the frontend?#5720 drift audit 6 May 2026Resolved 2026-06-03 — RESOLVED-IN-CODE — options live in a frontend constant FOCUS_ENCOUNTERS (History/Exam/Diagnosis/Management/Explanation/Consent), matching the spec; the empty conFocusOfEncounter:[] in config is the blank initial multi-select value, not a missing list. No DB seed needed. code:instances/ppp/src/constants/focus-encounter.ts.
Q15#5720 AUDIT — "Collated by representative" notice placement. Spec says QMF and MCR forms should show a "Collated by representative" notice. Is this a UI label that already exists in the rendered components, or is it missing from both forms?#5720 drift audit 6 May 2026Resolved 2026-06-09 — RESOLVED-IN-CODE — the representative-disclaimer notice is now present on both the MSF (code:instances/ppp/src/forms/MSFForm/MSFConsultantForm/FormMSF.tsx) and MCR (code:instances/ppp/src/forms/WBAConsultantForm/FormMCR.tsx) consultant forms; the wording is shared via buildRepresentativeDisclaimerLabel (code:instances/ppp/src/constants/wba-ratings.ts) and the MCR form gates Submit on the acknowledgement, mirroring MSF. The earlier tracked gap (#8174) is closed; see AC wba.mcr.representative-disclaimer-notice.
Q16#5720 AUDIT — MSF section name "goodMedPractice" vs spec grouping "Maintaining Practice". Confirm whether the section header label in the UI reads "Maintaining Practice" or "Good Medical Practice", and whether the 5 question texts match what's currently live.#5720 drift audit 6 May 2026Resolved 2026-06-03 — RESOLVED-IN-CODE — live UI header reads 'Maintaining Good Medical Practice'; the 5 question texts match the spec list (with two added parentheticals). code:instances/ppp/src/forms/MSFForm/MSFConsultantForm/FormMSF.ReviewSection.tsx, code:instances/ppp/src/constants/wba-ratings.ts.
Q23MCR / MSF / 360 participant model — does the shipped implementation match Matt's confirmed model? Matt confirmed (#8171, 2026-06-30, restated 2026-07-10): "MCR — assigned by user in form process — already built; MSF — yes Education Supervisor (now Named Professional Supervisor); 360 — either platform users or non platform users." Verify each against shipped code before citing.#8171 (Matt 2026-06-30 / 2026-07-10) †CONFIRMED-MATCHING-SHIPPED 2026-07-11 (#10368 code audit, per Matt #8171). All three participant models match the shipped implementation: MCR — user-assigned in the form process; mcrService (code:api/backend-ppp/src/services/cpd/mcrService/index.ts) with the super-admin-configurable collaboration window (systemSettings.collaborationWindowDays, default 7 business days, ?? 7 fallback — see Q20 / #9059) and all 9 capabilities gated by MCR_SHOW_OPTIONAL_CAPABILITIES (see AC wba.mcr.optional-capabilities-off-switch / #6998); form code:instances/ppp/src/forms/WBAConsultantForm/FormMCR.tsx. MSF — collated by the non-consultant supervisor (corrected 2026-10-03, Task 12: the shipped MSF form does NOT route to the supervisor. The assessor is a free choice of any user; see appraisals.msf.completed-by-professional-supervisor, drift #12755); code:instances/ppp/src/forms/MSFForm/MSFConsultantForm/FormMSF.tsx + FormMSF.config.ts with the three closing fields (see Q13 / #6999). The supervisor role label was the one unresolved sub-question. Q22 above now answers it (2026-10-03): it is a display rename to Professional Supervisor (#9703, built by #9718). 360 — either platform users or non-platform users; ppp360 router (code:api/backend-ppp/src/routers/procedures/appraisal-and-feedback/ppp360/index.ts), external raters via verify360ReviewerToken, minimum 7 raters + Educational Supervisor + Colleague Appraiser (see Q8 / #5009).
Q20 (b, c)MCR collaboration window: what happens when it closes, and does 360 have one? (Parts (b) and (c) of Q20; part (a) stays open.)transcript-2026-03-11 (~01:02:01); issue-7916 (2026-06-02); notion-appraisal-briefClosed 2026-10-03 (inferred from existing evidence, Rob approved). (b) The window's expiry is a due date. Nothing closes when it passes: the #7916 reminders carry on (an automated nudge at about 1 and 2 weeks, then the requester's "prompt again" after about 3 weeks), and the requester can extend the window. Matt's 2026-03-11 concern was that items don't remain outstanding, and his 2026-06-02 #7916 answer is the mechanism for that. Mismatch noted: the first automated reminder fires at 7 calendar days, before a 7-working-day window ends. (c) 360 is already time-limited by its 30-day magic-link expiry (REVIEW_360_TOKEN_TTL_DAYS env setting, default 30 days, token-helper.ts), which matches the Notion brief's "magic link that expires". It gets no separate window. External raters are not reminded today (drift #12756).††
Q22"Named Professional Supervisor" (NPS) — display-rename of the existing Educational Supervisor role, or a new distinct role/relationship? In restating the confirmed participant model (#8171, 2026-06-30) Matt noted the MSF supervisor is "yes Education Supervisor (now Named Professional Supervisor)". The spec already uses "Educational Supervisor" and "Professional Supervisor" interchangeably for the same non-consultant appraiser (see the Yearly Appraisal Lifecycle table and AC Professional Supervisor field enabled in MCR request form (#5088)). Two mutually-exclusive readings, do NOT infer (per the WBA-schema "reference, not requirements" rule): (1) NPS is a global display-label rename of the existing role — the underlying users.educationalSupervisorId FK / one(users) relationship and the persisted enum are unchanged, exactly the QMF→MSF precedent (#6999, where the display name changed but wbaType: 'qmf' stayed for data compatibility); (2) NPS is a new, distinct role with its own data relationship, additive to Educational Supervisor. Engineering strongly expects (1) given the interchangeable usage already in the spec and Matt's "now …" phrasing, but the choice determines whether a migration/new FK is needed. Schema note (do not act on): the shipped data model already carries two supervisor relationships — a role-named single FK users.educationalSupervisorId (educational_supervisor_id) and a generic many-supervisors join userSupervisors.supervisorId (supervisor_id, role-agnostic). If Matt picks (1), the label changes but the educational_supervisor_id column name is left stale-but-untouched (pure data-compat, QMF→MSF pattern); a physical column rename is a separate, non-urgent migration and is NOT implied by a display-rename. confirm-with-matt — do NOT rename any label, enum, or column, or add a role, until Matt confirms (1) vs (2).#8171 (Matt 2026-06-30) †ANSWERED — option (1), a platform-wide display rename (Matt, #9703, 2026-06-29; restated in the 2026-07-01 Dev Session). Matt filed #9703: "Vernacular change for non-consultant profiles - platform wide. Change 'Educational Supervisor' to 'Professional Supervisor'".† On 2026-07-01 Rob asked him to restate it, and Matt said: "I've decided to change educational supervisor to professional supervisor and nothing to change for the consultant user".† This was a rename of the existing role, not a new relationship. It is built: PR #9718 (2026-06-29) changed every user-facing label, and the educational_supervisor_id column was left as it was (QMF→MSF pattern). A read-only grep on 2026-10-03 found no user-facing "Educational Supervisor" string left in instances/ppp/src (stories and code comments excluded). The 2026-06-30 #8171 wording "Named Professional Supervisor" describes the same role (date corrected 2026-10-03: the comment is dated 2026-06-30; there is no 2026-07-10 comment on #8171). It is not a different label. Everything else in the MCR/MSF/360 participant model is confirmed-matching-shipped (see Closed Q23 below).
Q29CPD for the people who complete an MCR, MSF or 360: should they earn any? Matt's Task 12 brief asks to show the "accounting and logging of any triggers CPD hours/points" in CPD Home. Today only the subject earns CPD.#11210 (Matt 2026-07-29); notion-appraisal-briefClosed 2026-10-03: folded into Q27 (Rob approved). That completers should earn CPD is inferable: Matt's 2026-05-28 platform demo says senior users' supervisory and educational involvements "would then score against their CPD home", and his Task 4, 9 and 12 briefs (2026-07-29) ask for CPD "triggering if appropriate for both users". How much, and in which category, is the same open question as WBA sign-offs, so it is asked once as Q27. New ⏳ AC appraisals.lifecycle.completer-cpd-credit.††
Q30MCR final report: does the Professional Supervisor write one unified report? In 2026-03 Matt described an MCR as a report that "individually goes out to other people and then is unified by the educational supervisor". Today the MCR report view lists every consultant's responses side by side, with the Professional Supervisor first.transcript-2026-03-25 (~00:38:08); notion-appraisal-brief; transcript-2026-05-19 (~00:08:52)Closed 2026-10-03 (inferred from existing evidence, Rob approved): no. The side-by-side collated view is the final report. Matt's Notion brief asks for a final report of "multiple entries collated", and on 2026-05-19 he said the MCR submission is "already coded into the system … that shouldn't have really any adjustments". The 2026-03-25 "unified by the educational supervisor" wording is superseded by those later statements. Not checked: whether the Moreton Blue MCR report had a supervisor summary (no copy in the corpus).††
Original Vision ​

WBA Types ​

From walk-through notes (Jan 2026): "Four assessments — Descriptions of cases and..."

  1. CBD (Case-Based Discussion) — consultant reviews a case with the trainee
  2. CEX (Clinical Evaluation Exercise) — observed clinical encounter assessment
  3. DOPS (Direct Observation of Procedural Skills) — procedure skill assessment
  4. MSF (Multi-Source Feedback) — 360 feedback from multiple colleagues. Since #12836 an MSF is stored as an Appraisal record (type: 'appraisal', appraisalType: 'msf') and is no longer a WBA type; the Add WBA menu offers CBD, CEX, PBA and SSA only.
  5. 360 Review — broader multi-rater review (medical + non-medical + non-clinical)

WBA Form Schema (audit reference) ​

⚠️ This is a reference, not a requirements doc. Source: Matt's WBA Forms spreadsheet, May 2026. Matt iterated significantly during the Moreton Blue UAT pilot and the spreadsheet may not reflect those changes. Where existing forms in instances/ppp/src/forms/ differ from this schema, raise a Matt-confirmation question before changing code or seed data. Six form types (CEX, CBD, PBA, SSA, Quarterly MSF, MCR) share a common skeleton; differences are in the rating items and the optional "procedure" field on PBA/SSA.

Shared registrar (trainee) section — every WBA type:

FieldTypeNotes
DateAuto + editableDefaults to today
HospitalSelectQCH only for the pilot
URNTextUnidentified record number (case-based forms only: CEX/CBD/PBA/SSA)
SettingSelectOutpatient Clinic / Ward / Operating Theatre / Emergency Department
Case AcuitySelectElective / Trauma
Summary of CaseText
AssessorSelect consultantSearch-and-select consultant profile
Trainee Reflection ×4Text (optional)"Did well", "Learned", "Need to improve", "How will I achieve it"
Job DurationTextQuarterly MSF + MCR only
Post Graduate YearTextQuarterly MSF + MCR only (currently disabled in MCR/MSF — see #5095)
ProcedureSelect (tree)PBA + SSA only — picks from logbook procedure category tree

Shared consultant section — every WBA type:

  • FEEDBACK (free text + Yes/No): General, Strengths, Development needs, Recommended actions, Verbal feedback completed.
  • RATINGS scale: N Not assessed / D Development required / S Satisfactory / O Outstanding.
  • GLOBAL SUMMARY scale: Level 0–4 (PBA/SSA additionally split each level into a/b sub-levels — 9 total).

Type-specific consultant fields:

WBAExtra fieldRating items
CEXFocus of Encounter (multi-select: History / Exam / Diagnosis / Management / Explanation / Consent), Complexity (Levels 1–4)8 items: history, examination, diagnostic skills, management, judgement, communication, organisation, professionalism
CBD—9 items: medical record keeping, clinical assessment, diagnostic skills, management, judgement, communication, leadership, reflective practice, professionalism
PBA / SSADifficulty of Case (Easier / Average / More Difficult)10 items: indications & anatomy, consent, prep, analgesia, asepsis, technical aspects, unexpected events, documentation, communication, professionalism
Quarterly MSF"Collated by representative" notice16 items grouped: Clinical Care (5) / Maintaining Practice (5) / Learning & Teaching (2) / Relationships (4)
MCR"Collated by representative" notice7+ capabilities, each with Broad Feedback select + Specific Feedback text + Level 0–4 rating: professional values, professional skills, professional knowledge, health promotion, leadership & teamwork, patient safety, safeguarding (specialty extensions allowed)

Notes for engineering:

  • The "Procedure" field on PBA/SSA reuses the logbook procedure category tree (same source data as docs/features/logbook.md). One source of truth, two surfaces.
  • Per Matt 3 May, DOPS = SSA — same form, dual label in UI (#5718). Don't build a parallel DOPS schema.
  • The Quarterly MSF form ends with a free-text section asking for outstanding/development examples + a probity & health concerns toggle + an environments-observed multi-select (Primary care / Specialist / A&E / Critical care / MDT / Outpatients / Theatre / Ward / Other).
  • MCR is the collated consultant view — it is the natural surface for the "Professional Supervisor / Colleague Appraiser" role (Matt 3 May, Q9).
  • Implementation tracked in #5720.

Reviewer Groups (pending rename — #3958) ​

Current: generic reviewer group names Pending: rename to Medical Doctor / Medical Non-Doctor / Non-Medical†

WBA Charts ​

From walk-through notes (Jan 2026):

"WBA - each chart needs a legend showing what is required vs completed" "I.e. list of Principal Procedures — may need to be a bar chart"

Each competency chart must show required vs completed counts, with a visible legend.

All doughnut/pie widgets on the WBA page (e.g. WBA Actioned, MCR/MSF/360 Overview cards) use the canonical thin-doughnut config shared across PPP — innerRadius 75 / outerRadius 120 desktop, innerRadius 55 / outerRadius 90 mobile (cutout ~62%). This keeps the ring thickness visually consistent with CPD, Surgical Logbook, Appraisals, Timesheets and Retrospective widgets (PR #7831).

Yearly Appraisal Lifecycle ​

Confirmed 2026-05-03 (Matt). Three phases only — no separate "locked" state.

PhaseDescriptionWho acts
DraftDoctor creates and edits the appraisalDoctor
Review with appraiserSubmitted to Colleague Appraiser (consultant) or Professional Supervisor (non-consultant)Appraiser reviews
Signed offAppraiser approves — record is locked, no further editsAppraiser

Engineering note: the previously proposed 4th "locked" state is dropped. Signed off implicitly locks the record. The existing appraisalWorkflow.createDraft and appraisalWorkflow.submit tRPC mutations (#5007) cover phases 1→2. Phase 3 is covered by appraisalWorkflow.complete / completeAppraisal, which records appraiserSignedOffAt — no new mutation needed.

Appraiser side (#12992-#12994, Task 12B): the appraiser reaches the appraisal from their To-do (type yearlyAppraisal, from the appraisal_submitted trigger), the notification, or the "Appraisals to review" button on Appraisals, at /appraisals/review (queue) and /appraisals/review?id=N (review screen: Begin review, section comments, PDP comment, endorse goal, Defer with reason, Sign off with outcome and summary). The doctor's annual appraisal form shows the review state (Review with appraiser / Signed off / Deferred), the appraiser's comments and the outcome, and is read-only once submitted or signed off. Known gap: the form still does not persist the section content itself (#12995), so a shared section can be empty on the appraiser's screen.

Yearly Appraisal — Annual CPD-Plan Review × Appraisal × PDP Crossover (#8178) ​

Source: Dev Session transcript 2026-06-03 (~00:11:40–00:15:40), raised by Matt and spun out as #8178. The Yearly Appraisal is not yet built. Matt described it as a crossover element linking CPD Home + Appraisal + Roadmap — the annual CPD-plan review (a CPD Home compliance step) rolls into the yearly appraisal and the creation of a Professional Development Plan (PDP / SMART goals), which is then discussed with the doctor's Colleague Appraiser (consultant) / Professional Supervisor (non-consultant; called Educational Supervisor before #9703).

Sequence Matt described (the "two separate things" model):

StepWhat happensWhere it livesLogged as
1Annual CPD-plan review — review last year's CPD plan + create a new one (the CPD Home compliance step)CPD HomeThe annual-review log
2That rolls into the yearly appraisal + PDP / SMART-goal creation, discussed with the appraiserAppraisalsThe appraisal/PDP log

Matt expected these to log as two separate things (the CPD-plan annual review first, then the appraisal/PDP), not one combined record.† The intent is iterative — to replace external appraisal systems over time.†

Roadmap crossover — "linked, not merged" (ppp-roadmap Q10 / #8167). SMART goals (annual reflection / PDP) live on the Roadmap year-round and evolve freely; the yearly Appraisal imports the currently-active goals as a point-in-time PDP section at review time, and exports any supervisor-endorsed goals back to the Roadmap on sign-off. This keeps the signed Appraisal a legally stable locked archive while preserving organic goal-setting. Reuses ppp_smart_goals + import/export logic in appraisalWorkflowService — no new table.† See ppp-roadmap.md Q10 and cpd-home.md for the annual-review compliance step.

Engineering note: this section is the spec enabler (#9091) for build slices #9092 (data model + state machine), #9093 (annual CPD-plan review compliance step in CPD Home), #9094 (appraisal record + PDP/SMART-goal creation + appraiser discussion), and #9095 (Roadmap crossover pull-in/push-back). The lifecycle states are the three phases confirmed 2026-05-03 (above); this crossover model adds the CPD-Home→Appraisal roll-in and the Roadmap import/export, both Matt-sourced.

Build status — #9092 shipped the foundational data model + state machine: the appraisal_annual_cpd_reviews table (the annual CPD-plan review log, one per doctor per year) with a not_started → in_progress → completed → rolled_into_appraisal lifecycle, plus an appraisals.annual_cpd_review_id linkage column that records the "two separate things, linked (not merged)" relationship. The user-facing CPD-Home review step (#9093), the appraisal/PDP creation + appraiser discussion (#9094), and the Roadmap import/export (#9095) build on this model. The appraisals.yearly.* ACs below stay ⏳ until those user-facing slices land — #9092 delivers only the persistence + transition layer they depend on.

Yearly Appraisal → Appraisals (rename — #3956) ​

Current: tabbed form for "Yearly Appraisal"; left-hand menu item reads "Yearly Appraisal" Pending: rename left-hand menu item to Appraisals (plural)†; replace tabbed interface with progress-indicator modal dialogue showing step completion indicators†

Consultant Guidance ​

From walk-through notes (Jan 2026):

"Need to create guidance for consultant to help identify what they need to do on each screen — in the app"

In-app contextual guidance for consultants performing WBA reviews.

Action Icon Change (#3957) ​

Replace eye icon with three-dots/arrow on WBA action buttons — reserve eye icon strictly for "view only" actions.†

Matt's 17 Apr 2026 Notion Brief — Split Intent ​

Matt now treats Appraisal, WBAs, and Yearly Appraisal as three distinct features. Content lives in this combined spec until the Wave 6 file-rename lands.

Appraisal (source: Notion) ​

  • Issue: collection is administratively difficult, multiple systems, no digital repository across training levels.
  • Purpose: easy, collatable, exportable; track involvements to drive performance.
  • Functions:
    • MSFs — one consultant responsible for collating multiple sources of feedback (departmental meeting for non-consultants; Colleague Appraiser path).
    • MCRs — multiple consultants write reports about non-consultants; the multi-consultant report selector lists consultant profiles only.†
    • 360 — feedback from doctors and non-doctors (all users); collection outside the platform.
    • Yearly Appraisal — see sub-section below.

Work Based Assessments (source: Notion) ​

  • Issue: WBAs are onerous; engagement is low above the mandated minimum.
  • Purpose: fully digital process satisfying current requirements + automated CPD scoring + track assessor/educator involvement + capture critical-assessment progression (e.g. Critical Condition WBAs — CBDs / CEXs per speciality; or Principal Procedure Level-4 competencies per user learning objectives).†
  • Functions: PBAs, SSAs (= DOPS†), CEXs (Clinical Examinations), CBDs (Case-Based Discussions). Competency scores collected and mapped on dashboard.
  • Outstanding task: spreadsheets defining current web-forms to be added (@Rob or ops).

Yearly Appraisal (source: Notion) ​

Status: yet to be built.

Completed with the Colleague Appraiser (for Consultants) or the Educational Supervisor (for Non-Consultants).† Non-consultants have two supervisor types: Educational Supervisor (appraisals) and Clinical Supervisor (practical work). (Brief dated 21 Apr. Educational Supervisor was renamed Professional Supervisor platform-wide on 2026-06-29, #9703; see Q22.)

  • Summary of Qualifications to date.
  • Review of Registration (+/- Indemnity).
  • Summation of CPD activities for the year; PDP plans and completion of goals — the yearly appraisal summarises CPD and PDP, rolling over career goals and prior-year data including long-term career aspirations.†
  • Surgical cases summary (from Logbook).
  • Optional: WBAs / Retrospective / other Appraisal outputs.
  • Where-to for the following year — goals and steps to completion.
  • Long-term aspirations.

Acceptance Criteria ​

WBA Creation & Submission ​

  • ⏳ CBD, CEX, MSF WBA types available
  • ✅ Surface DOPS = SSA equivalence in UI (alias/label, not a separate type) — Matt 3 May
  • ⏳ SSA and PBA have separate review forms (#5025)
  • ⏳ 360 Review supports multi-rater groups
  • ✅ WBA draft from logbook labelled 'draft' not 'submitted'†
  • ✅ MSF required fields validated before submit
  • ✅ WBA required fields validated before submit
  • ✅ Assessor Autocomplete selection persisted to saved record (#5022) — CreateViewUpdateForm.tsx keeps the selected assessor outside the search results; BDD @issue-5022
  • ✅ WBA page does not spam 'Access denied' on 60s poll (#5023) — a trainee's assessor notification now goes through notifications.createFromClient; BDD @issue-5023
  • ✅ wba.form.sticky-save-mobile — on a phone, the WBA / Reflection / Feedback form's primary action bar (Save / Submit / Verify) stays pinned and reachable above the on-screen keyboard while editing the last fields. The WBA create/review dialogs, the linked reflection/WBA/feedback modals on a logbook case, and the Work-Queue record dialog open with scroll="paper" (bounded scroll container) and the WBA, Reflection, Give-Feedback and Received-Feedback forms pass sticky on their primary FormLayout.Actions row. Implements the shared design-system.form-actions.sticky-mobile behaviour for the retrospective forms.†‡ (#9324)
  • ✅ wba.appraisal.date-prompt — a doctor sets their appraisal date from a lightweight "When is your appraisal?" prompt (a date; an optional appraiser is not built yet, #12879), opened by the onboarding step "Set your appraisal date" (/appraisals/upcoming?setAppraisalDate=1) or the Appraisals page header button. appraisalWorkflow.setAppraisalDate stores it in the periodEnd of the caller's appraisal for that date's year (created as a draft if none, moved in place while still a draft, refused once submitted). No schema change: periodEnd is what the Roadmap already reads as the appraisal meeting. The full 8-step Appraisal Form is unchanged.†‡ (#12787)
  • ✅ wba.form.auto-populate-from-logbook — WBA form auto-populates case security and principal procedure from linked logbook entry.† (#6353)
  • ✅ wba.reflection.no-duplicate — system prevents adding a second reflection to a logbook event that already has one.†
  • ✅ wba.form.three-sequential-boxes — WBA feedback displays in three sequential boxes (Feedback → Ratings → Global Summary); each unlocks only after the prior box's question is answered.† (#6355)
  • ✅ wba.form.single-fire-save — saving (and submitting) a WBA, a Reflection, or a Feedback form disables the primary action and shows a pending state until the create/update mutation settles; a rapid double-tap fires the mutation at most once, so a mobile double-tap can no longer create a duplicate WBA / reflection / feedback record. Implemented via the shared SubmitButton primitive (see design-system.md → design-system.submit-button.single-fire). † (#9323)
  • ✅ wba.cex.rating-items-source-wording — CEX rating items 6 and 7 read "Communication and listening skills" and "Organisation and time management" (reverted to the CEX source sheet); no near-duplicate management label with item 4.† (#6995)
  • ✅ wba.msf.closing-sign-off-fields — the MSF form ends with three closing fields: a specific-examples free text, a probity & health concerns Yes/No toggle with a conditional "if yes, please state" detail, and an environments-observed multi-select (Primary care / Specialist practice / A&E/admissions / Critical care / MDT / Outpatients / Theatre / Ward / Other).† (#6999)
  • ✅ wba.msf.display-name — user-facing copy reads "MSF" / "Multi Source Feedback" (not "QMF" / "Quarterly Multi Source Feedback"); MSF is stored as an Appraisal record (type: 'appraisal', appraisalType: 'msf'), no longer a WBA subtype (#12836).† (#6999)
  • ✅ wba.ssapba.nine-level-global-rating — the PBA and SSA global/overall rating picker offers the 9-level a/b scale (Level 0, 1a, 1b, 2a, 2b, 3a, 3b, 4a, 4b); CEX/CBD and MCR keep their 5-level scales; existing stored 5-level values still display on old records.† (#7001)
  • ✅ wba.mcr.optional-capabilities-off-switch — MCR Part 1 keeps all 9 capability areas by default; items 8 (Education and training) and 9 (Research and scholarship) can be hidden via a single config off-switch for AU enterprises that opt out, without hard-removing them.† (#6998)
  • ✅ wba.mcr.representative-disclaimer-notice — both the MSF and MCR consultant forms show the "Collated by representative" disclaimer notice (the consultant confirms the feedback represents consultation with multiple colleagues who observed the named subject); the wording comes from one shared builder and the MCR form gates Submit on the acknowledgement, mirroring MSF.†‡ (#8174)
  • ✅ wba.add.labelled-action-with-typed-menu — WBA add control is a clearly-labelled "Add WBA" button opening a menu of full assessment-type names, each resolving to a working create form.†‡ (#7806)
  • ✅ wba.header.acronyms-spelled-out — the WBA page explains its acronyms in the platform: the page title reads "Work Based Assessments" in full, not just "WBA", and the add menu lists the full assessment-type names.†‡ (#11194)
  • ✅ wba.form.condensed-density — the WBA create/edit form uses compact brand-density spacing (design tokens, no hardcoded gaps) so it is fast to scan and fill with less scrolling.†‡ (#7817)
  • ✅ wba.close.two-button-discard-confirm — closing the WBA modal with unsaved changes shows exactly two clear choices ("Discard changes" / "Keep editing"), not three (Cancel/No/Yes).†‡ (#7817)
  • ✅ wba.list.shared-section-cards — Submitted and Received list panels render inside the shared SectionCard primitive.†‡ (#7828)
  • ✅ wba.status-chip.tokenised-palette — the WBA list status chip uses SEMANTIC_COLORS.status.* tokens (no inline hex), mirroring the MCR fix.†‡ (#7828)
  • ✅ wba.cbd-modal.equal-section-spacing — sibling section blocks in the CBD create modal share the same compact top spacing (no extra space above/below any block).†‡ (#7828)
  • ✅ wba.filters.pill-only-date — /wba renders exactly one date-range control: the shared PillDateRangeFilter inside the filter bar; the legacy in-header range on <WidgetSection> is removed.†‡ (#7834)
  • ✅ wba.list.responsive-column-priority — the WBA list DataGrid adopts the DSDG responsive column-priority model (hideBelow, see design-system.md → design-system.table.grid-overflow-and-density): the Hospital and Assessor/Sender columns drop below the md breakpoint instead of being crushed, and the Type column keeps a 180px minWidth so its chip never wraps. Part of the wave-2 per-table adoption sweep — the WBA slice of #9893.†‡ (#9893)
  • ✅ wba.ssa.procedure-field-matches-pba — the SSA (DOPS) Add-WBA form renders the same Surgery/Procedure lookup (Principal Procedure / Treatment Acuity / Broad Area / Specific Area) as PBA, matching the "PBA + SSA only" Procedure field row in the shared registrar-section table above and the "same form, dual label in UI" DOPS=SSA directive (#5718). WBA_PROCEDURE_ASSESSMENT and WBA_TYPE_CATEGORY_MAP both classify SSA as a procedure assessment alongside PBA.†‡ (#12701)
  • ✅ wba.cbd.review-form-no-cex-fields — the CBD assessor review form (Received WBA(s) tab → Review) does not render CEX-only fields; it starts directly with the shared Critical Condition + Feedback sections, matching CBD's "no extra field" row in the type-specific-consultant-fields table below.†‡ (#12702)

Cadence & Scheduling ​

  • ✅ wba.cadence.meter-distributes-roadmap — Cadence Meter sets WBA frequency for non-consultants and distributes required assessments across the roadmap.†
  • ✅ wba.cadence.set-during-onboarding — the WBA/appraisal cadence (CBD, CEX, PBA, SSA, MCR, 360, MSF) is collected in the unified Cadence Settings card (alongside the surgical targets) via the CadenceTracker component with a selectable 3/6/12-month period. The cadence drives two engines: the appraisal subset (MSF, MCR, 360) feeds the reminder engine (processAppraisalNotifications → calculateAppraisalDueDates), which distributes the cadence across the doctor's selected cadencePeriodMonths (#8118) rather than assuming a fixed 12-month plan; the WBA-assessment subset (CBD, CEX, PBA, SSA) feeds the roadmap distribution (milestone-engine, see wba.cadence.meter-distributes-roadmap, which also honours the period). See onboarding.md onboarding.cadence.unified-step.†† (#7344, #8041, #8118; code:instances/ppp/src/modules/MyProfile/UserForm/index.tsx CadenceTracker, code:api/backend-ppp/src/services/scheduledNotificationService/index.ts calculateAppraisalDueDates, code:api/backend-ppp/src/services/roadmap/milestone-engine.ts)
  • ✅ wba.cadence.default-values — the WBA/appraisal cadence comes pre-filled with Matt's defaults, limited to the types the doctor's role grid allows (wba.config.role-based-grid): one of each WBA type (CBD/CEX/PBA/SSA) per 3 months; MCR twice a year, due at months 3 and 9 (both due-date engines place MCRs mid-interval, #12730; offered to non-consultants only, #12806); MSF twice a year, due at months 6 and 12; 360 once a year. The doctor confirms or changes these in the onboarding cadence step (wba.cadence.set-during-onboarding); nothing is applied without that confirm step.††† Built (#12717): instances/ppp/src/utils/cadenceDefaults.ts (resolveInitialCadence, used by useUserFormLogic.ts) pre-fills these defaults, limited to the types the role grid allows; the non-surgeon onboarding step confirmWbaCadence (#12729, onboardingService) makes the doctor confirm them (cadenceDefaults.test.ts, onboardingService/index.test.ts).
  • ✅ appraisals-wba.runner.completion-nudge — a seasonal appraisal "runner" completion-nudge reminds doctors to finish their annual appraisal 3 / 2 / 1 months before the completion-due date. The schedule is DB-backed and NOT hardcoded — the completion-due anchor (season_due_month / season_due_day) and the relative-to-due lead offsets (lead_months) live in the appraisal_reminder_settings row, read at run time by the run-appraisal-nudges Container Apps Job. The Dec–Jan window is the seeded DEFAULT (complete by 31 January, so the final 1-month nudge lands in December), configurable per region with no code change. Each milestone fires once per cycle via a scheduledKey idempotency stamp in the notification metadata. Distinct from wba.cadence.set-during-onboarding, which paces the per-type MSF/MCR/360 reminders.†‡ (#9650) — engineering DEFAULT shipped; Dec–Jan default pending Matt's async confirm, for-discussion.

CPD Scoring ​

  • ✅ wba.cpd.consultant-sign-off-credit — when a consultant completes a WBA sign-off (CBD/CEX/PBA/SSA moves to verified), CPD is credited automatically to that signing consultant, so their teaching/assessor involvement counts toward their own CPD.††† Each sign-off earns a fixed 0.5 h under Teaching/Supervision, within the yearly supervision cap (Rob, #12712, resolves Q27 for WBA sign-off). Verified built (5 Oct 2026, Task 9 wave 2): cpdUserActivityHandler.ts awardCompleterCredit writes a Provided WBA assessment row of exactly 0.5 h for the signing consultant (asserted per sign-off in cpdUserActivityHandler.integration.test.ts), and the live BDD scenario "A consultant who signs off a WBA is credited 0.5 CPD hours" (#11194, @issue-11194) signs a WBA off through activities.signOffWba and reads the new row back from cpdUserActivities.getAll.† activities.signOffWba now triggers the CPD handler (it had stopped when PR #12570 moved sign-off off the generic update path), so the trainee's "WBA assessed by consultant" credit fires again too. (#12712)

Security & Audit Integrity ​

  • ✅ assessorId filter enforced server-side on getByFilter
  • ✅ Backend blocks trainee mutation of WBA in pending review (#5054)
  • ✅ wba.security.update-owner-only-signoff-by-assessor — through the activities router, only a WBA's owner (or a super-admin) may edit it, and the owner only while it is a draft (#5054). Through activities.create a WBA is created only as draft or pending, and activities.update only writes draft / pending to it. A non-owner can never change a record's userId, assessorId or cpdSupervisorId; nobody can change a record's type into or out of wba. The owner may withdraw a pending (shared, not yet signed off) WBA back to a draft, the one write the guard accepts on a shared WBA (#12822): the assessor immediately loses access and their to-do is cleared, and the owner can then correct and re-share it; an in-place edit of a pending WBA stays refused. The assigned assessor signs off a pending WBA only through activities.signOffWba (never the owner, even if they named themselves as assessor). It writes the status, the competency_signoff audit row and the to-do clear in one transaction (the sign-off fails if the audit write fails) and notifies the trainee server-side. The assessor may read a WBA once it is shared (pending or later) via getById, getByFilter and getAll; an unshared draft stays owner-only. Known gaps at this commit, tracked separately: the MCR router (api/backend-ppp/src/routers/procedures/cpd/mcr) is a second, unguarded WBA path — mcr.create can store an MCR (type = 'wba', wbaType = 'mcr') already verified, mcr.update can set any MCR to verified, mcr.getAll / mcr.getById / mcr.extendCollaborationWindow are unscoped, and the owner can move a draft WBA onto that path by changing its wbaType to mcr (all #12586); non-WBA types are still editable by non-owners (#12584).†‡ (#12568, #12543, #12544)
  • ✅ activities.getByType returns only the caller's own activities (super-admin unscoped) — a signed-in user can never list another user's WBAs or their decrypted patient identifiers by type (#12585)†
  • ⏳ BDD covers cross-user assessor tRPC scenarios (#5262)

Appraisal Form & Workflow ​

  • ✅ appraisal.form.progress-indicator-modal-flow — the annual appraisal opens as a guided progress-indicator modal flow, not a free-jump tabbed interface: the modal shows an overall completion meter (a determinate progress bar labelled "Appraisal completion progress" with an "N of M sections complete" summary) and the step navigator renders a completion tick against each finished section. Advancing through the sections (Next, or jumping forward via the step navigator) marks the sections passed as complete and advances the meter; a successful Submit fills the meter to 100%.††‡ (#9654, #3956)
  • ✅ Save Draft and Submit call real tRPC mutations
  • ⏳ Personal Details fields populate from profile (#5093)
  • ⏳ "Appraisal Form" button carries a tooltip: "Open your annual appraisal".†
  • ⏳ "Request Appraisals" button carries a tooltip: "Request an MCR, 360, or MSF peer assessment".†
  • ⏳ Request Appraisals dropdown opens the selected type form (#5092)
  • ⏳ Close on step 7 dismisses the form, not opens MCR modal (#5091)
  • ⏳ UI exists for assigning an appraiser before submit (#5010)
  • ⏳ Post Graduate Year field enabled in MCR/MSF request forms (#5095)
  • ⏳ Professional Supervisor field enabled in MCR request form (#5088)
  • ⏳ MSF records categorised under MSF tab, not 360 (#5094)
  • ✅ Appraiser dropdown filter — Consultant appraisee → Colleague Appraiser (Consultants only); Non-Consultant appraisee → Professional Supervisor (Consultant or Non-Con) — these are synonymous per Matt 3 May
  • ✅ MSF/360 individual rater identities are visible to appraisee — Matt 3 May: not anonymous (was previously specced as a bug)
  • ✅ 360 review enforces minimum 7 reviewers — PPP360SubmitInputSchema superRefine in inputSchema.ts rejects submission when reviewers.length < PPP_360_MIN_REVIEWERS (=7); integration tests assert exactly 7 reviewers required (#5009). Open question: the "plus mandatory Professional Supervisor (non-consultant) and Colleague Appraiser (consultant)" role composition check is NOT yet enforced server-side — the code validates reviewer count only, not the two named roles.
  • ✅ appraisals.mcr.on-brand-shared-template — the MCR appraisal page uses the shared PPP page template: a PageHeader title + labelled actions and SectionCard bordered panels, with no hardcoded hex/px (status chips + radii from design tokens); all 9 MCR capabilities and functionality preserved.†‡ (#7825)

MSF / MCR / 360 Lifecycle (Task 12 brief, #11210) ​

Matt's Task 12 brief asks for the lifecycle of each appraisal type, for both non-consultant and consultant initiators, in four steps: (1) the user starts it and selects the people involved, (2) those people (internal or external) complete it, with in-app and email notices and reminders, (3) it returns to the initiator as a final report, (4) any CPD it triggers is logged and shown in CPD Home.† The Yearly Appraisal ACs are in the Yearly Appraisal sections below (Task 1). The ACs here were checked read-only against the code on 2026-10-03.

  • ✅ appraisals.lifecycle.user-initiated — the doctor starts every MSF, MCR and 360 from "+ Request" on the Appraisals screen and picks the people involved in the request form. There is no admin or department scheduling step.† (This restates Q1, answered by Matt on 2026-05-03: user-initiated, department scheduling deferred.)‡
  • ✅ appraisals.mcr.non-consultant-requester-only — only a non-consultant can request an MCR. Multiple consultants then write reports about that non-consultant. A consultant receives and completes MCRs but never requests one.††† Built in #12754 (PR #12781): the "+ Request" picker no longer offers MCR to a consultant, a consultant's /appraisals/mcr?add=1 deep link no longer opens the request dialog, and mcr.create rejects a consultant requester with FORBIDDEN. The MCR form's consultant branch (Colleague Appraiser) is still in the code but no longer reachable.‡
  • ✅ appraisals.mcr.many-consultants-selected — in the MCR request the non-consultant selects several consultants, not just their supervisors. Each one writes their own report.†† The MCR becomes verified once every selected consultant has responded.‡
  • ✅ appraisals.mcr.collaboration-window — each MCR has a collaboration window for the consultants to respond. It defaults to 7 business days and is settings-driven (systemSettings.collaborationWindowDays), so it can be changed without a code change. The old 48 hours was too tight for a report started on a Friday.†††‡ The 7-business-day value is Rob's acting-PO default; Matt has only said "extend". Working vs calendar days is Q20. The expiry is a due date, not a cut-off: nothing closes when it passes, the #7916 reminders carry on (appraisals.reviewer-non-response.*), and the requester can extend the window (Q20 (b), closed 2026-10-03). Mismatch: the first automated reminder fires at 7 calendar days, before a 7-working-day window ends. The report view shows the expiry ("Collaboration window closes on …") and offers the requester a once-per-report Extend window button (#12984, #12985). Working vs calendar days, and what should happen when the window closes, are Q20. Nothing shows or enforces the expiry today.†
  • ✅ appraisals.msf.completed-by-professional-supervisor — an MSF is completed by one consultant, who collates feedback from several sources (for example at a departmental meeting). For a non-consultant that is their Professional Supervisor (formerly Educational Supervisor). For a consultant it is their Colleague Appraiser.†††† Built in #12755: the MSF request form pre-fills the Assessor with that named person and offers no one else; with none on the profile it points the doctor at My Profile. activities.create / update refuse an MSF assessor who is not the requester's named Professional Supervisor / Colleague Appraiser (FORBIDDEN). Whether a free choice should remain as an override is open for Matt. This corrects Q23's 2026-07-11 "matching shipped" resolution for MSF.‡
  • ✅ appraisals.360.platform-and-external-raters — 360 raters can be anyone, doctors and non-doctors, platform users or not. Each rater gets an email with a magic link that expires, so non-platform raters can respond without an account (the "email lite" version). The request needs at least 7 raters (Q8). Each rater is grouped as Medical Doctor / Medical Non-Doctor / Non-Medical. The 30-day link expiry (REVIEW_360_TOKEN_TTL_DAYS env setting, default 30 days, token-helper.ts) is the 360's response window; 360 has no separate collaboration window (Q20 (c), closed 2026-10-03). External raters are not reminded today (drift #12756, see appraisals.lifecycle.reviewers-notified-and-reminded).†††‡
  • ✅ appraisals.360.final-review-by-appraiser — once the raters have responded, the 360 goes for final review by the requester's Colleague Appraiser (consultant) or Professional Supervisor (non-consultant), who writes the summary the requester then sees.†‡ The supervisor is a separate sign-off reviewer, not one of the 7 raters. The backend accepts a 360 with no supervisor (supervisorId is nullish). Whether the supervisor and Colleague Appraiser must also be raters (Q8) is not enforced; see the 360 AC under Appraisal Form & Workflow.
  • ✅ appraisals.lifecycle.reviewers-notified-and-reminded — everyone asked to complete an MSF, MCR or 360 is told about it: platform users in app, non-platform users by email. A reviewer who doesn't respond is reminded the same way (the two-reminder rule, appraisals.reviewer-non-response.*).†† Drift (#12756): the automated reminders now run daily on the NAS backend (APPRAISAL_REVIEW_REMINDER_TICK, 22:00 UTC, registered in entrypoints/api-container.ts), and external 360 raters are now reminded by email, with a platform-user rater also told in app (#12838). MCR request notices are created server-side (mcr.feature).‡ (MSF is notified in app and by email: notifyAssessorOnWbaSubmission.)
  • ✅ appraisals.lifecycle.returns-to-requester-with-final-report — when the reviewers have finished, the requester is told (closed-loop, e.g. "your MCR has been completed by the consultant") and can open a final report that brings the multiple entries together.††† Built as follows. MCR: a notice for each consultant's review, and a report view with every consultant's responses side by side, Professional Supervisor first. There is no single "MCR complete" notice. MSF: a sign-off notice and the supervisor's report. 360: a notice after supervisor sign-off, a grid of every rater's answers and the supervisor's summary.‡ Whether an MCR also needs one unified report written by the Professional Supervisor is Q30. For MCR the side-by-side collated view is the final report; no unified supervisor report is needed (Q30, closed 2026-10-03).
  • ✅ appraisals.lifecycle.cpd-credit-in-cpd-home — a completed MCR, MSF or 360 credits CPD to its subject automatically, as AMC "Multisource feedback using a structured framework" (1 per activity, no maximum), and the entry shows in CPD Home. The rule lives in cpd-automations.md § Auto-scoring — multisource feedback.†‡ CPD for the people who complete them is appraisals.lifecycle.completer-cpd-credit below; the amount is Q27. CPD for the people who complete them (consultants, supervisors, raters) is not specified; that is Q29.
  • ✅ appraisals.lifecycle.completer-cpd-credit — a platform user who completes an MCR, MSF or 360 for a colleague (a consultant writing an MCR report, the supervisor collating an MSF, a platform-user 360 rater or the 360 final reviewer) has that involvement credited to their own CPD automatically, and it shows in their CPD Home. How much it earns, and in which category, is Q27 (shared with WBA sign-offs). Built (Rob, #12712; Q27 amount and category still formally with Matt): each completer earns a fixed 0.5 h Supervision then Teaching, within the yearly caps, for WBA sign-off, MCR, MSF, the 360 final reviewer, and every 360 rater who is a platform user (matched by email) and completed their questionnaire, once the 360 is signed off (#12986).††† (#11210)
  • ⏳ appraisals.consultant-workload-acknowledgement — deferred to v2. When a consultant has a lot of appraisals submitted to them and not yet actioned, PPP shows them a supportive prompt. It acknowledges the volume ("we appreciate you are completing/have received a lot this month") and notes that this work is recorded, so they can show their administrative workload to their employer. The aim is completion and accountability without overloading consultants.† Not built. No prompt text or trigger exists in the code. Deferred (v2), 2026-10-03: on 2026-05-19 Matt called it a possible "version two" item and "a ticket on hold", so the unstated threshold ("a lot this month") is not asked now. The threshold ("a lot this month") is not stated; the build should ask Matt for it.†

Evidence Pack & Export ​

  • ✅ /appraisals/evidence-pack route renders the evidence pack export panel, not redirects to /mcr
  • ✅ Evidence pack PDF export downloads correct file
  • ✅ Evidence pack export implemented for AHPRA/GMC submission — consolidated PDF (personal details, CPD, MSF/360, WBA, mandatory training, PDP, appraiser comments, outcome)

Notifications ​

  • ✅ Email sent to WBA assessor on submission (#5026) — activitiesWorkflow.ts buildWbaAssessorEmail, best-effort; BDD @issue-5026. UAT mail delivery depends on the mail configuration (#12803, ops), not code
  • ✅ MSF / 360 notification click opens correct dialog
  • ✅ WBA audit notification deep-links to record
  • ✅ 360 supervisor review notification routed to work queue
  • ✅ appraisals.reviewer-non-response.two-automated-reminders — a non-responding reviewer gets two automated reminders: the first at ~1 week, the second at ~2 weeks; PPP stops automating after the second.†‡ (#7916) Drift (#12756, found 2026-10-03): the reminder logic existed but was never scheduled; it now runs daily on the NAS backend (registerScheduledJob(APPRAISAL_REVIEW_REMINDER_TICK, …) in entrypoints/api-container.ts; scheduler-lambda.ts is the disabled Azure path). It now covers external 360 raters too (#12838, by email).
  • ✅ appraisals.reviewer-non-response.requester-prompt-after-3-weeks — after the two automated reminders (~3 weeks) the requester sees a "prompt again" control noting the reviewer has had two automated reminders and may be on leave / could be contacted directly; no "PPP is not responsible" disclaimer.†‡ (#7916)
  • ✅ appraisals.reviewer-non-response.friendly-reminder-notification-only — the requester's manual prompt sends a friendly-reminder notification ("Just a friendly reminder there's an outstanding {item} for you to action") to the reviewer; notifications only, creates no additional action item.†‡ (#7916)

Charts & Visualisation ​

  • ⏳ Each competency chart has legend showing required vs completed. Decision needed (not built): #12716 deliberately removed the pooled target because curricula set level 4 per condition or procedure, so there is no single "required" figure to put in a legend. The PBA widget now shows N / M per index procedure where the curriculum states a count; either retire this AC or confirm the legend should be per condition or procedure (Matt).
  • ⚠️ Bar chart for Principal Procedures: required vs completed — the PBA competency widget now shows per-index-procedure level-4 bars (wba-index-procedure-progress), with a required count only where the curriculum states one; the logbook-side Principal Procedures summary remains surgical-case based.
  • ✅ Charts clickable to drill into case details
  • ✅ wba.widget.hash-colour-submitted — WBA widget uses hash colours for submitted-not-complete; solid colours for complete.† (#6357)
  • ⚠️ wba.competency.preloaded-horizontal-bars — competency and critical-condition widgets pre-populate all items for the doctor's specialty before any data is entered (never empty for a specialty that has curriculum data). The horizontal bars track progress toward the level-4 requirement of the doctor's specialty curriculum: per critical condition (level-4 CBD/CEX evidence) and per index procedure (the curriculum's level-4 PBA count). This is not a fixed count of 3.†††† CONTRADICT (both versions kept): the earlier version of this AC read "horizontal bars track progress toward 3 × level-4 assessments" (transcript 2026-05-19 §16). Matt's later statements (Q19, the same session; #9123 2026-06-28; transcript 2026-07-08) make the target per-specialty and curriculum-derived. Built (#12716): the global constant is gone. The WBA dashboard datasets return the specialty curriculum's own level-4 wording (curriculumRequirement: critical-condition certification for CEX/CBD, index-procedure certification for PBA). The Critical Conditions widget tracks level 4 per condition ("N / M conditions", M being the number of the specialty's conditions, a Level 4 chip on each condition that has one). The competency widgets show the level-4 count with the curriculum's words, because curricula set level 4 per condition / per procedure, not as one pooled count per WBA type. A specialty with no curriculum data held (#12713) shows its items with "Your specialty's level-4 requirement isn't available yet", never an invented target. SSA has no curriculum level-4 requirement, so its widget shows the level-4 count with no requirement note. Built (Task 9 wave 2, #11194): the PBA widget now also lists every curriculum index procedure with its own level-4 count (getWBACompetencyData returns indexProcedures, attributing each PBA through its logbook procedure's principal procedure, whose names are the curriculum's index-procedure names; WBACompetencyWidget renders the rows). A target N / M shows only where the curriculum states an explicit level-4 PBA count (parseRequiredLevel4PbaCount), never an invented one. A PBA on a procedure with no principal procedure counts toward the pooled total only. Still open: the exact per-condition rule is pending Matt (Q24), so this stays ⚠️. ‡ (#6358, #12716)
  • ⏳ wba.dashboard.competency-long-term-trend — the WBA feature dashboard shows the doctor's competency stage increasing over time, so long-term progression is visible, not just the current level distribution.†† The visual form is Q25. (#11194)
  • ⏳ wba.widget.cadence-goals — the WBA page shows a goals widget tied to the cadence meter (as on Appraisals): how many WBAs to complete, and how many of those fall in the current 3-month period.† Position and display (horizontal runner, top or bottom) are Q26. A WBASummaryWidget ("progress against N-month cadence targets") exists in components/widgets/WBAWidgets/ but is not rendered on /wba. Confirm it fits this AC before reusing it. (#11194)
  • ✅ wba.config.role-based-grid — role-based configuration grid defines which WBA forms, statuses, and critical conditions apply to consultants vs non-consultants.† (#6359) — packages-shared/ppp/src/constants/wba-role-config.ts is the single source of truth (WBA_ROLE_CONFIG_GRID); consultants get MCR/MSF, non-consultants get CBD/CEX/PBA/SSA + MSF and specialty critical-condition + level-4 tracking.

Reviewer Groups & UI ​

  • ✅ 360 reviewer groups renamed Medical Doctor / Medical Non-Doctor / Non-Medical†
  • ✅ Admin not selectable in reviewer dropdowns
  • ✅ WBA primary action icon: three-dots or arrow, not eye†
  • ⏳ Eye icon reserved for view-only actions†

Yearly Appraisal / Appraisals Module ​

  • ✅ Module renamed from 'Yearly Appraisal' to 'Appraisals'†
  • ⏳ Tabbed layout replaced with progress-indicator modal dialogue.†
  • ⏳ In-app consultant guidance on each WBA review screen
  • ✅ appraisals.yearly.appraiser-by-user-type — appraiser defaults by user type: Consultant → Colleague Appraiser; Non-Consultant → Professional Supervisor (appraisals) and Clinical Supervisor (practical work).††† CONTRADICT (both versions kept): the 2026-04-23 wording said Educational Supervisor. Matt renamed the role to Professional Supervisor on 2026-06-29 (#9703) and restated the rename on 2026-07-01 (Q22). The label rename is built (#9718). The default-by-user-type behaviour is built (SummarySection.tsx defaults the appraiser by userType; consultant case tested in SummarySection.test.tsx).
  • ✅ Yearly appraisal rolls prior year goals and career aspirations to new cycle (built: appraisalPdpService.rollPriorYearGoals, wired by useAppraisalData and PdpSection).†
  • ✅ appraisals.yearly.data-sharing-opt-in — the doctor controls which yearly-appraisal elements are shared with their appraiser; sharing defaults to WITHHOLD / opt-in (private by default). The annual appraisal form has a Data Sharing step listing a per-element switch (Personal Details, Year in Review, Well-being, CPD Activities, Multi-Source Feedback & competency, PDP), each rendered OFF; nothing is shared with the appraiser until the doctor explicitly opts an element in — the safe default for clinical PII (opt-in cannot accidentally over-share). Matt confirmed "Option A" (#9649, 2026-06-29). The form model defaults every element to withheld and the filterSharedSections seam omits every withheld element, so an un-opted appraisal yields an empty share. Scope of #9649: the doctor-facing control + private-by-default form model + the share filter. End-to-end backend persistence of the preferences and appraiser-read enforcement are now BUILT (#9764): the per-element choices persist to appraisals.sharing_preferences via appraisalWorkflow.persistSharingPreferences (owner-only, draft-only, wired into Save Draft / Submit), and the appraiser read path (getAppraisalSummary) filters server-side through filterAppraisalForAppraiser so the appraiser only ever sees the elements the doctor opted to share — withhold-by-default, never trusted to the client. The doctor (owner) always sees their own full appraisal; only the appraiser-only read is filtered; identity is derived from ctx (no IDOR). for-discussion retained for Matt's async re-confirm (auto-populate-convenience tension noted).†††‡ (#9649, #9764)
  • ⏳ Appraisal runner prompt at screen top nudges completion in December and January.†
  • ⏳ Report/widget date picker defaults to full calendar year; user can override to specific time window.†

Yearly Appraisal — CPD-Review × PDP Crossover (#8178) ​

Scaffold ACs for the spec enabler (#9091). All ⏳ not-built; each is Matt-sourced (transcript 2026-06-03 / resolved ppp-roadmap Q10). See § Yearly Appraisal — Annual CPD-Plan Review × Appraisal × PDP Crossover.

  • ✅ appraisals.yearly.annual-cpd-review-compliance-step — CPD Home presents an annual CPD-plan review compliance step: the doctor reviews last year's CPD plan and creates a new one; completing it logs the annual-review step. The step renders as a RequirementTile in the CPD Home wheel sidebar and drives the S1 forward-only state machine not_started → in_progress (start) → completed (complete); start/complete each write an annual_cpd_review_start / annual_cpd_review_complete audit entry.†‡ (#9093)
  • ✅ appraisals.yearly.review-rolls-into-appraisal — completing the annual CPD-plan review rolls into the yearly appraisal and the creation of a PDP / SMART goals.†‡ (#9093, #9094)
  • ✅ appraisals.yearly.logs-as-two-separate-things — the annual CPD-plan review and the appraisal/PDP are recorded as two separate logged items, not one combined record.†‡ (#9092, #9093, #9094)
  • ✅ appraisals.yearly.pdp-discussed-with-appraiser — the PDP / SMART goals created in the appraisal are discussed with the doctor's Colleague Appraiser (consultant) or Professional Supervisor (non-consultant).†‡ (#9094; appraiser UI #12992)
  • ✅ appraisals.yearly.roadmap-crossover.import-active-goals — at review time the yearly appraisal imports the currently-active SMART goals from the Roadmap as a point-in-time PDP section.†‡ (#9095)
  • ✅ appraisals.yearly.roadmap-crossover.export-endorsed-goals — on appraiser sign-off the appraisal exports any supervisor-endorsed goals back to the Roadmap; the signed appraisal stays a locked archive (linked, not merged).†‡ (#9095; endorse and sign-off UI #12992)

Yearly Appraisal — year in review + PDP (28 Jun 2026 brief) ​

Source: Matt's 28 Jun update of the Notion page "PPP Road Map / Yearly Appraisal / Compliance Runner ie What's Up Next", which he walked through in the 2026-07-01 Dev Session (~00:48:49–00:55:22). The task issue is #11144. These ACs extend the #8178 crossover above and do not replace it.

  • ⚠️ appraisals.yearly.draws-in-year-data — the yearly appraisal is the doctor's year in review. It draws in the year's CPD Home activity; clinical involvements (surgical cases, WBAs, feedbacks, reflections, audits); appraisals (MCR, MSF, 360); the PPP Passport; and, optionally, PPP Pulse (wellness). The doctor does not re-enter any of it. Each element stays subject to appraisals.yearly.data-sharing-opt-in.†† Drift: appraisalAggregationService.getAggregatedAppraisalData collects the profile, CPD plan, wellness, WBA summary, 360, previous appraisals and passport mandatory training. It does not collect surgical cases, feedbacks, reflections, audits or MCR/MSF. getWbaSummaryForAppraisal now counts SSA alongside CEX/CBD/PBA (#12741 partial); the other sources above are still not drawn in.
  • ⏳ appraisals.yearly.pdp-three-parts — after the review, the doctor builds next year's Professional Development Plan in three parts: (a) next year's CPD plan (the CPD Home crossover, appraisals.yearly.annual-cpd-review-compliance-step), (b) a long-term plan (e.g. becoming a consultant), and (c) steps for next year tied to the long-term plan, such as publications, presentations, exams, conferences and courses.†† The shipped PDP items (appraisal_pdp_items) are a single generic SMART-style list with no long-term-plan / next-year-step split. The field design waits on the L2P reference (Q28).
  • ⏳ appraisals.yearly.pdp-cadence-catch-up — if the doctor has fallen behind their Roadmap cadence, the next-year steps recommend raising the cadence for the following year to get back on track (Roadmap crossover).††

Appraisal Widgets ​

  • ⏳ MSF has separate "received and reviewed" and "submitted and completed" widgets.†
  • ⏳ MCR has its own distinct widget, separate from MSF.†
  • ⚠️ appraisals.widgets.per-role-matrix — the MSF, MCR and 360 widgets show received and reviewed and submitted and completed according to the user's role: MSF: consultant both, non-consultant submitted/completed. MCR: consultant received/reviewed only, non-consultant submitted/completed. 360: both roles get both, as two widgets (received/reviewed, submitted/completed).†† Built (#12757): each role gets the panels in the matrix, headed "Submitted and completed" / "Received and reviewed", and Completed shows for every role. Still drifting: a 360 "received" count is always 0, because the dashboard dataset does not track 360 reviewers as assessors (the code comment in appraisalsReceived.ts says so).‡ The Notion brief also asks for a higher-fidelity display than the wheel ("?HOW BEST TO DISPLAY"); that design choice is still open.
  • ✅ appraisals.widgets.cadence-meter — Horizontal cadence meter at top of appraisals screen shows upcoming task timeline; renders MSF/MCR/360 due dates as uniform circular milestone markers on a 12-month rail with a today needle, derived from the user's cadenceTracker and cadencePeriodMonths settings via the #8118 cadence engine. Each marker is a clickable control that navigates to the matching appraisal tab (MCR / 360 / MSF); its accessible name is exactly its visible label (MCR / 360 / MSF) with the due-date exposed as a tooltip description, not folded into the name (#7773). All markers share one diameter (consistent-size circles, not variable-width chips) and the rail/needle stroke is the lighter mid-weight per Rob's 2026-07-08 roadmap feedback (#10188).††‡ (#9652, #10188)
    • ✅ The rail reserves enough height for the tallest same-month marker stack, so markers never overlap the month-label row or render outside the card's border — confirmed via live getBoundingClientRect() measurement at 1440px and 375px (viewport-independent root cause).†‡ (#12714)
    • ✅ Each month's marker stack is centred on that month's column (directly above its label), so a December milestone sits inside the timeline rather than on the card's right edge; markers are 24px so the widest label ("MSF") fits inside the circle unclipped. Measured at 1440px and 375px.†‡ (#12737)
    • ✅ The month labels use the opaque text.secondary token (gray[600], measured 6.32:1 on the card), so they meet WCAG AA 4.5:1 and all twelve render in one tint. The previous text.disabled (black at 38% alpha) measured 2.68:1, and the translucent glyphs anti-aliased warmer at some sub-pixel offsets, which is why Mar and Jul looked a different colour. No month is meant to be highlighted.†‡ (#12738)
  • ⏳ Platform-wide widget states use "received and reviewed" / "submitted and completed" vocabulary.†

Reflections ​

  • ✅ appraisals.reflections.tag-to-logbook-entry — a standalone reflection can be tagged to a specific surgical logbook entry, creating a traceable link to the case; once tagged, the linked case is echoed back as a read-only "Linked to: …" confirmation so the link is visible at a glance. Does NOT include the Pendleton-model field structure (separate concept — see Q21).†‡ (#6399, #10134)

Audit Events ​

  • ⏳ Creating an audit event notifies all case participants automatically (see Q18).†
Status ​
Sub-featureStatusNotes
CBD, CEX, MSF WBAs✅ BuiltWorking
DOPS WBA❌ Missing#5024 — absent from platform; required for UK GMC
SSA / PBA consultant review⚠️ Bug#5025 — shared form; domains not separated
360 Review✅ Built#3579 blank dialog fixed; anonymisation missing #5008
360 minimum reviewers✅ Done#5009 — count enforcement shipped (inputSchema.ts superRefine, PPP_360_MIN_REVIEWERS=7); role composition (Supervisor + Colleague Appraiser) not yet enforced
WBA validation (save/submit)✅ Built#5021, #5089
WBA assessor persistence✅ Built#5022 — Autocomplete persisted (BDD @issue-5022)
WBA polling error✅ Built#5023 — assessor notification via createFromClient (BDD @issue-5023)
WBA server-side assessor ownership✅ Built#5027; follow-up BDD #5262
WBA trainee mutation guard⚠️ Bug#5054 — no server-side block during review
WBA draft from logbook📋 Specified#3950 on-hold
Appraisal Save Draft / Submit✅ Built#5007 — real tRPC mutations
Appraisal Personal Details pre-populate⚠️ Bug#5093 — fields blank
Request Appraisals dropdown⚠️ Bug#5092 — always opens generic form
Appraisal step-7 Close⚠️ Bug#5091 — opens MCR modal
Appraiser assignment UI❌ Missing#5010 — backend blocks; no UI path
MCR / MSF Post Graduate Year field⚠️ Bug#5095 — disabled, no explanation
MCR Professional Supervisor field⚠️ Bug#5088 — permanently disabled; blocks submission
MSF tab categorisation⚠️ Bug#5094 — MSF records show under 360
Colleague Appraiser dropdown filter⚠️ Bug#5063 — includes CPD-only users
Evidence pack route✅ Done#5096, #5011 — renders the evidence pack export panel
Evidence pack PDF export✅ Done#5011 — server-side @pdfme/generator PDF, all sections
WBA assessor email notification✅ Built#5026 — assessor email on submission (UAT mail config #12803)
Competency charts with required/completed⚠️ MissingNo legend — no issue yet
Reviewer group rename📋 Specified#3958 on-hold
Admin excluded from reviewers📋 Specified#3959 on-hold
Action icon change📋 Specified#3957 on-hold
Appraisals rename📋 Specified#3956 on-hold
Consultant in-app guidance📋 SpecifiedNo issue yet
Process flow documentation📋 Specified#3961 on-hold
Yearly appraisal goal rollover❌ Not startedTranscript 2026-05-19 §6
Yearly appraisal draws in the year's data⚠️ DriftMatt 2026-06-28 / 2026-07-01: no surgical cases, feedback, reflections, audits or MCR/MSF; SSA uncounted — #12741
Yearly appraisal PDP structure📋 SpecifiedMatt 2026-06-28 / 2026-07-01: three-part PDP + cadence catch-up; L2P reference pending (Q28)
User-controlled data privacy✅ Done#9649 control + private-by-default; persist + appraiser-read enforcement #9764
Appraisal runner (Dec–Jan)❌ Not startedTranscript 2026-05-19 §8
MSF / MCR separate widgets❌ Not startedTranscript 2026-05-19 §9
Cadence meter on appraisals screen✅ Done#9652 — AppraisalCadenceMeter component
WBA hash colour widget states✅ DoneTranscript 2026-05-19 §15 (#6357)
Pre-loaded competency widgets✅ DoneTranscript 2026-05-19 §16 (#6358)
Level-4 competency horizontal bar✅ DonePer-specialty curriculum target, tracked per condition (#12716); exact rule pending Q24
WBA cadence defaults⚠️ DriftMatt 2026-09-29 defaults (#11144) not pre-filled — #12717
WBA stepwise three-box web forms⚠️ UnverifiedTranscript 2026-05-19 §22 — confirm against live forms
Reflections — free-text (not Pendleton)✅ ResolvedMatt D8178.8 (2026-07-13, #9653): free-text per original spreadsheet; Pendleton = Feedback tool only. Shipped form already matches (Q21)
Reflections — tag to log entry❌ Not startedTranscript 2026-05-19 §19
Audit event auto-notification❌ Not startedTranscript 2026-05-19 §23 (see Q18)
WBA auto-populate from logbook✅ DoneTranscript 2026-03-25 §WBADataPull
WBA reflection deduplication❌ Not startedTranscript 2026-03-25 §ReflectionDedup
Appraiser defaults by user type❌ Not startedTranscript 2026-04-23 §AppraiserRoles
WBA cadence meter for non-consultants❌ Not startedTranscript 2026-04-23 §CadenceMeter
Role-based WBA config grid✅ Done (#6359)Transcript 2026-05-19 §RoleConfigTable
Date picker defaults to full year❌ Not startedTranscript 2026-05-19 §DatePickerDefault

Open issues on label:feature:appraisals-wba (6)

#Title
#12741fix(appraisals-wba): yearly appraisal does not draw in surgical cases, feedback, reflections, audits or MCR/MSF
#12756fix(appraisals-wba): reviewer non-response reminders never run, and external 360 raters are never reminded
#12838fix(appraisals-wba): external 360 raters are never reminded; 360 platform raters and MCR requests get no server-side in-app notice
#12865feat(appraisals-wba): deferred to v2. When a consultant has a lot of appraisal…
#12866feat(appraisals-wba): after the review, the doctor builds next year's **Profession…
#12867feat(appraisals-wba): if the doctor has fallen behind their Roadmap cadence, t…

Fetched at: 2026-10-13T07:29:49Z — live list on GitHub.

Change Log ​
DateChangePR/IssueReason
2026-10-06Yearly Appraisal Save Draft and Submit now persist each section the doctor entered (personal details, Year in Review, well-being, CPD, MSF and competency, summary notes) plus a server-computed snapshot of the in-year aggregates (CPD activity count, well-being questionnaire stats, 360 and WBA counts) into appraisal_sections through the new owner-only, draft-only appraisalWorkflow.saveSectionContent. Before this only the PDP goals and the sharing preferences were saved, so the appraiser review screen showed empty sections. The appraiser read path is unchanged and still withholds every element the doctor did not opt in; the summary and mandatory_training sections are never shareable.#12995Task 12 audit finding #12995: the appraiser had nothing to review (Refs #11210).
2026-10-06Task 12 wave 2: the MCR report view shows the collaboration window expiry and offers the requester a once-only Extend window button; platform-user 360 raters earn the 0.5 h completer credit; the Appraisals widget counts 360s received; an admin can override the appraisal completion-nudge cron. Spec statuses corrected: reviewer reminders (external 360 raters, #12838), completer credit, the appraiser default and the two-reminder rule are built; pdp-discussed-with-appraiser and export-endorsed-goals are backend only until the appraiser review UI exists. The yearly lifecycle scenarios with no step definitions are parked @wip (#12946).#12984, #12985, #12986, #12987, #12991, #12946, #12838, #9703, #127125 Oct 2026 Task 12 reachability check (Refs #11210).
2026-10-06Yearly Appraisal appraiser side built: an Appraisals to review queue and review screen at /appraisals/review (begin review, section and PDP comments, endorse goals, defer with reason, sign off with outcome and summary), a yearly appraisal To-do item and appraisal_submitted trigger and notification routing, and the doctor's form now shows the review state, appraiser comments and outcome and locks once submitted or signed off. The pdp-discussed-with-appraiser and export-endorsed-goals ACs now have a UI.#12992, #12993, #12994, #12995Task 12 reachability check found the appraiser procedures had no client caller (Refs #11210). Section content persistence is still open as #12995.
2026-10-05The annual appraisal form now calls the in-year aggregation, pre-filling CPD activity count, wellbeing average (rescaled to /10) and 360 response count. Appraiser by user type and prior-year goal rollover are marked built. Surgical cases, reflections, audits and MCR collectors remain to be added.#11192, #12959Task 8 reachability check of 5 Oct 2026 (items 18, 19, 20).
2026-10-05Task 9 wave 2: an owner can withdraw a shared (pending) WBA back to a draft (server accepts only that write on a shared WBA; assessor loses access and to-do); the PBA competency widget lists every curriculum index procedure with its own level-4 count; critical-conditions counts exclude drafts; the consultant sign-off 0.5 h CPD credit is verified in code and covered by a live BDD scenario and a per-sign-off integration assertion, so its AC is now built. Stale spec lines corrected: cadence defaults ship (#12717, #12729), assessor persistence (#5022), no Access denied spam (#5023), assessor email (#5026), MSF and 360 are Appraisal records not WBA types (#12836).#11194, #12822, #12712, #12717, #12729, #12796, #12836, #5022, #5023, #50265 Oct 2026 Task 9 reachability check (Refs #11194): several ACs and status rows still said not-built or drift for behaviour that already shipped; the reachability report was wrong that an owner could already pull a pending WBA back (the #5054 service guard blocked it).
2026-10-05Task 12 audit fixes: seeded supervisor and consultant personas open the appraisal screens with a pre-filled Assessor; a 360 reviewer notice is aggregated per 360 so later raters no longer 500; a co-assessor sees an MCR on their To-do; MCR, MSF and 360 completers earn the fixed 0.5 h Teaching/Supervision credit; Evidence Pack is linked from Appraisals; the dated runner sits at the top of Appraisals and no longer says 'this week' for overdue items; the annual appraisal defaults the appraiser to the Professional Supervisor or Colleague Appraiser; suggested cadence copy matches the rows a role sees.#12897, #12898, #12899, #12900, #12901, #12902, #12903, #9703, #127125 Oct 2026 Task 12 audit (Refs #11210) and Rob's 4 Oct decision on #12712.
2026-10-04The mobile Received WBA(s) list rows now show the sender and a visible Review (pending) or View (reviewed) button, matching desktop.#12582Issue #12582: #12544 fixed only the desktop grid; mobile rows had an unlabelled eye icon and no sender.
2026-10-04MCR due dates now fall halfway through each interval of the cadence period (2 a year → end of month 3 and month 9) in both calculateAppraisalDueDates engines (backend notifications and the frontend cadence meter). MSF (forward) and 360 (backward) are unchanged.#12730Issue #12730: Matt (#11144, 2026-09-29) asked for MCRs at months 3 and 9; the engine counted them backward from period end (months 6 and 12). Rob decided the change applies to every user (4 Oct). Existing users' MCR reminder dates move accordingly.
2026-10-04The yearly-appraisal WBA summary (getWbaSummaryForAppraisal) now counts SSA alongside CEX/CBD/PBA, adds ssaCount, and includes it in totalCount and the evidence pack text.#12741Issue #12741: SSA is a live WBA type but was silently dropped from the year-in-review count. Surgical cases, feedbacks, reflections, audits and MCR/MSF are still not drawn in, so appraisals.yearly.draws-in-year-data stays ⚠️ and #12741 stays open.
2026-10-04Appraisal reviewer non-response reminders now run daily at 22:00 UTC on the NAS backend (registerScheduledJob(APPRAISAL_REVIEW_REMINDER_TICK, …) in api-container.ts). The ~7 / ~14 day cadence and two-reminder cap are unchanged.#12756Issue #12756: checkAndCreateAppraisalReviewReminders was reachable only through the triggerReminders mutation, which nothing calls, so no reviewer was reminded. External 360 rater reminders, the requester prompt reaching them, in-app notice for platform 360 raters and server-side MCR notices remain open (split out).
2026-10-04Added AC wba.appraisal.date-prompt. New appraisalWorkflow.setAppraisalDate stores the doctor's appraisal date in appraisals.periodEnd (create draft, move in place while draft, refuse once submitted), opened from a lightweight dialog.#12787, #11189The onboarding step opened the Appraisal Form, which has no date input. Rob decided on 2026-10-04 on a lightweight prompt; no schema change.
2026-10-04The WBA form's Hospital field now uses the shared instance-wide hospital search that the Logbook and Reflection forms use. Typing any word of a hospital name finds it, the doctor's own hospitals are listed first, and GB trusts such as "Royal Berkshire NHS Foundation Trust" can be picked instead of an Australia-only list. The selected hospital resolves by id.#12796Bug #12796 (SMAN demo practice run, 3 Oct 2026). The WBA Hospital list was filtered to the profile country, so a doctor could not find their own hospital for a CBD. Not done yet - the MSF, MCR, Give Feedback and Timesheet pickers still use the country-filtered list, and the duplicate hospital rows (canonical name plus aliases) are still in the data.
2026-10-04WBA Case Acuity now also offers Acute / Emergency, Inpatient and Outpatient / Chronic. The critical-condition question is hidden when the specialty has no critical conditions configured, instead of an unanswerable Yes/No plus an administrator alert.#12799Issue #12799: medical and acute presentations had only Trauma/Elective, and specialties with no configured list showed a dead-end question. Seeding critical conditions for IMT/orthopaedics needs curriculum data and is left open.
2026-10-04The Appraisals page no longer injects startDate/endDate into its Grid child (passDateRangeToChild={false}), removing the React unknown-prop warning on load.#12805Issue #12805: WidgetSection cloned its date range onto a DOM element, logging a console error at every load.
2026-10-04The suggested default cadence no longer offers an MCR count to consultants. MSF and 360 defaults apply to every role; CBD/CEX/PBA/SSA stay gated by the role grid.#12806Issue #12806: Rob decided (4 Oct) MCR defaults are non-consultant only (as #12754) while MSF and 360 apply to every role.
2026-10-04A 360 rater who is also a platform user is now notified in app when the 360 request is sent, in addition to the email. External-rater email reminders and the requester's "prompt again" by email stay open.#12838, #11144#12838, from #11210 (Task 12 brief 2026-07-29): "secondary user/s (internal or external) completing (also please show how these users are notified/reminded ...)". AC appraisals.lifecycle.reviewers-notified-and-reminded stays drift.
2026-10-04Signing off a WBA now credits CPD again. The assessing consultant earns a fixed 0.5 h under Supervision (Teaching once Supervision is at its yearly cap), and the trainee's "WBA assessed by consultant" credit fires. wba.cpd.consultant-sign-off-credit flips to built.#12712, #11194Rob's decision on #12712 (2026-10-04) fixed the amount and category. signOffWba (PR #12570) had stopped calling the CPD handler, so neither user was credited. A logbook-inherited operative time no longer overrides a WBA's fixed credit.
2026-10-04The MSF, MCR and 360 overview widgets follow Matt's per-role matrix. MSF - consultant both, non-consultant submitted. MCR - consultant received only, non-consultant submitted. 360 - every role gets both panels. Panels are headed "Submitted and completed" / "Received and reviewed", and the Completed count shows for every role. 360 received stays 0 until 360 assessors are tracked.#12757, #11194Spec drift #12757 against the Notion Appraisal brief and the 2026-05-19 Dev Session. AC appraisals.widgets.per-role-matrix stays marked drifting because of the 360 received dataset gap.
2026-10-04MSF (Multi-Source Feedback) is now an Appraisal record, not a WBA. It is stored on activities as type = 'appraisal' with appraisal_type = 'msf' (new column, new appraisal activity type); the old type = 'wba' / wba_type = 'qmf' rows are backfilled by migration with their ids unchanged, so review tasks, notifications and deep links keep resolving (a pre-remodel ?wbaType=qmf MSF link still opens the dialog). Every surface follows - Action Items to-do tiles read "MSF (Appraisal)" and are never "WBA (QMF)", the work-queue Type column and MSF filter, review URL and dialog routing, the assessor sign-off (activities.signOffWba, same two-party draft/share/sign-off rules as a WBA), the MSF list and form, WBA lists and counts (MSF drops out), the Appraisals summary/received widgets, roadmap milestones, the multi-feature report, reviewer reminders and the requester prompt, and the MSF CPD credit.#12836Rob, 4 Oct 2026 (#12836): full remodel rather than the label-only fix. MSF is conceptually an Appraisal instrument, and grouping it under WBA confused trainees and consultants. The migration swaps the activity_types enum because Postgres cannot use a newly added label in the transaction that adds it, and drizzle runs every pending migration in one transaction.
2026-10-04External 360 raters (no platform account) who have not responded are emailed a fresh magic link at about day 7 and day 14 by the same scheduled reminder job that notifies in-app reviewers, and the requester's "Prompt again" now reaches them by email too (it was in-app only and only reached the 360 supervisor). A 360 rater whose email matches a platform account is reminded and prompted in app instead. findPendingAppraisals returns both kinds of rater. The owner's manual resend, the scheduled reminders and "Prompt again" share one email builder.#12838Issue #12838 (split from #12756; source #11210 Task 12 and the Notion Appraisal brief, "email and magic link"). Reminders previously covered only MCR assessors, the MSF assessor and the 360 supervisor, so an external rater who never answered was never nudged. Reminders already sent are read back from last_reminder_sent_at so no migration is needed. The in-app notice for platform raters and the server-side MCR request notification were already built.
2026-10-04After "Confirm cadence", the Cadence Tracker holds the confirmed note's height only until the next edit in the card (a Period change or a count edit), then collapses to the "Cadence saved" line.#12752Bug #12752, from the #12731 design re-review. The reserved height stayed forever, so at 375px "Cadence saved" sat above about 130px of blank space (measured 132px). The Period selector still does not move at the moment of confirm.
2026-10-04appraisals.msf.completed-by-professional-supervisor moves from drift to built. The MSF Assessor is pre-filled with the requester's Professional Supervisor (non-consultant) or Colleague Appraiser (consultant) and offers no one else. With none set, the form points to My Profile. activities.create and activities.update refuse any other MSF assessor with FORBIDDEN.#12755Matt on #8171 (2026-06-30), "MSF - yes Education Supervisor (now Named Professional Supervisor)", and the Notion Appraisal brief ("Colleague Appraiser" for consultants). The assessor was a free choice of any user. Whether a free choice should stay as an override is left open for Matt.
2026-10-04A consultant's Appraisals page no longer offers an MCR. The "+ Request Appraisals" guidance reads "Request a 360 or MSF peer assessment", the overview subtitle drops MCR, and the cadence empty state promises only 360 and MSF milestones. Both header buttons keep their visible label as the accessible name, with the guidance on title instead of an injected aria-label.#12826Bug #12826, found by the PR #12781 design review. After #12754 a consultant can't request an MCR, but the page copy still offered one, and MUI's Tooltip put a diverging aria-label on the button (CLAUDE.md #7773). The MCR tab and widget stay, because a consultant still receives and reviews MCRs.
2026-10-03appraisals.mcr.non-consultant-requester-only moves from drift to built. The "+ Request" picker hides MCR from consultants, a consultant's /appraisals/mcr?add=1 deep link no longer opens the request dialog, and mcr.create rejects a consultant requester with FORBIDDEN.#12781, #12754Matt (#11210) — "MCR (non consultant only)". The drift was recorded by the Task 12 spec refresh (#12820); PR #12781 builds it. The MCR form's consultant branch is left in place but is no longer reachable.
2026-10-03Cadence Meter rail now reserves enough height for the tallest same-month marker stack — milestone markers no longer overlap the month-label row or render outside the card's rounded border. Confirmed via live getBoundingClientRect() measurement at 1440px and 375px.#12715, #12714Task 9 follow-up audit (#11194) of the Cadence Meter area PR #12703 never tested found the rail container's height was derived only from fixed padding, independent of marker-stack size — a viewport-independent layout bug, not a mobile-only cosmetic issue.
2026-10-03Corrected Q27 and wba.cpd.consultant-sign-off-credit, which said logbook sign-off had the same consultant-CPD gap as WBA. It doesn't — logbook already credits the consultant (built, but drifting from Matt's #9200 rules, #12725); the gap is WBA-specific (the CPD handler's WBA branch covers only MCR/QMF).#12712, #12725The Task 4 spec-refresh lane checked the code (cpdUserActivityHandler) and found the logbook credit built; the cross-feature claim had been carried over from a reconciliation pass without a code check.
2026-10-03Phase C open questions resolved from existing evidence (Rob approved 2026-10-03). Q22 moved to Closed with its citation date corrected from 2026-07-10 to 2026-06-30 (the #8171 comment is dated 2026-06-30). Q30 closed (the side-by-side collated MCR view is the final report) and its sentence struck from appraisals.lifecycle.returns-to-requester-with-final-report. Q20 parts (b) and (c) closed as a separate Closed row: the collaboration-window expiry is a due date, nothing closes, the #7916 reminders continue and the requester can extend (with the mismatch noted that the first reminder at 7 calendar days falls before a 7-working-day window ends); 360 uses its 30-day magic-link expiry. appraisals.mcr.collaboration-window and appraisals.360.platform-and-external-raters updated. Q20 narrowed to part (a), working vs calendar days. Q29 folded into Q27, which is now the single question on the CPD amount for WBA sign-offs and MCR/MSF/360 completers; new ⏳ AC appraisals.lifecycle.completer-cpd-credit. appraisals.consultant-workload-acknowledgement marked deferred (v2), citing Matt's "ticket on hold" of 2026-05-19.#11210, #9059, #7916, #12712, #12756Research on 2026-10-03 against #8171, #9059, #7916, #11183, #11210, Matt's Appraisal Notion page and the 2026-03-11, 2026-05-19 and 2026-05-28 transcripts. Spec-only; no BDD scenarios added, per the spec-refresh rule.
2026-10-03The appraisal ACs were refreshed for Task 1 (the WBA ACs were already done in Task 9 and are untouched). Q22 is answered as a platform-wide display rename, Educational → Professional Supervisor (Matt #9703, 2026-06-29; restated 2026-07-01), and it is already built (#9718). The appraiser-defaults AC is rewritten as appraisals.yearly.appraiser-by-user-type, keeping the old wording struck through. Three new Yearly Appraisal ACs come from Matt's 28 Jun Notion brief and the 2026-07-01 walkthrough. draws-in-year-data is marked drift (#12741). pdp-three-parts and pdp-cadence-catch-up are not built yet. New Open Question Q28 asks for the L2P appraisal reference Matt said he provided.#11144, #9703, #12741Async Task spec refresh, Task 1 (#11144). The newest dated statement from Matt wins. This also resolves the #12720 handoff "Task 12 / appraisals-wba: Educational Supervisor → Professional Supervisor" (2026-07-01 Dev Session).
2026-10-03The MSF/MCR/360 ACs were refreshed for Task 12. The WBA (Task 9) and Yearly Appraisal (Task 1) ACs are untouched. A new section, "MSF / MCR / 360 Lifecycle", has 11 ACs mapped to Matt's four-step #11210 brief, sourced from the Notion Appraisal brief, Matt on #8171 (2026-06-29/30) and the Dev Sessions. Four are marked drift. MCR is requestable by consultants (#12754). The MSF assessor is a free choice, not the Professional Supervisor or Colleague Appraiser (#12755). Reviewer reminders never run and never reach external 360 raters (#12756), which also flips appraisals.reviewer-non-response.two-automated-reminders from ✅ to ⚠️. The widgets don't follow the per-role matrix (#12757). Q20 is rewritten as partly answered. Q23's MSF resolution is corrected, with the old wording struck through. New Open Questions Q29 (CPD for the people who complete MCR/MSF/360) and Q30 (unified MCR report).#11210, #12754, #12755, #12756, #12757Async Task spec refresh, Task 12 (#11210). The newest dated statement from Matt wins. Read-only code checks found that Q23's 2026-07-11 "matching shipped" audit was wrong for MSF, and that the ✅ on the two-reminder AC rested on the logic existing, not on it running.
2026-10-03Task 9 top-up. Re-checked the WBA ACs against meeting tabs the pilot missed. These were the Transcript tabs for 16 Sep, 22 Sep, 24 Mar, 7 Apr and the 28 May Platform Demo, the 29 Sep Full notes, and every other Transcript tab that matches WBA terms. Added wba.header.acronyms-spelled-out (built). Added the 2026-05-28 demo as the earliest source for wba.cpd.consultant-sign-off-credit and wba.dashboard.competency-long-term-trend; it agrees with the later brief and supersedes nothing. Added the 29 Sep Full notes to wba.cadence.default-values. Q3 now records Matt's 2026-07-14 refinement (admin console, separate "Training sign-off requirements" tab) and that the data was still incomplete on 2026-07-29 and 2026-09-29. Q25 and Q27 have 2026-05-28 context but stay open.#11194, #12712, #12713The pilot read each Drive doc's first tab only. The 16 Sep, 22 Sep, 24 Mar and 7 Apr Transcript tabs hold no WBA requirements. In the 24 Mar transcript, "CBD" is a mis-transcription of "CPD". No newer source supersedes anything the pilot wrote, and no new drift was found.
2026-10-03Spec refreshed against Matt's newest dated statements (Notion WBA brief, Dev Sessions 2026-05-19 to 2026-09-29, Matt on #9123/#11144/#11194). wba.competency.preloaded-horizontal-bars rewritten to a per-specialty curriculum level-4 target and marked drift (was a fixed "3 × level-4"; both versions kept). Added wba.cadence.default-values (drift), wba.cpd.consultant-sign-off-credit, wba.dashboard.competency-long-term-trend and wba.widget.cadence-goals. Open Questions Q24–Q27 added.#11194, #12712, #12716, #12717Async Task spec refresh (Task 9 pilot). The newest dated Matt statement wins. Code drift is filed as bugs #12716 (fixed level-4 target) and #12717 (no cadence defaults); the unbuilt asks are tracked under #11194 and #12712.
2026-10-03wba.competency.preloaded-horizontal-bars partly built (stays ⚠️): the fixed REQUIRED_LEVEL_4_ASSESSMENTS: 3 is removed. The Critical Conditions widget tracks level 4 per condition out of the specialty's conditions; the competency widgets show the level-4 count with the specialty curriculum's own wording; a specialty with no curriculum data says the requirement isn't available yet, and SSA (no curriculum level-4 requirement) shows no note. Per-index-procedure PBA tracking is not built.#12716Bug #12716 — Matt's newest statements (Q19 2026-05-19, #9123 2026-06-28, Dev Session 2026-07-08) make the level-4 target per-specialty and curriculum-derived. No number is invented where curriculum data is missing (#12713); the exact per-condition rule stays Open Question Q24.
2026-10-03Cadence Meter markers are centred on their month column (above the month label) instead of the column's right edge, and are 24px rather than 20px. December milestones now sit inside the timeline and the "MSF" label is no longer clipped.#12737Bug #12737 — end-of-month anchoring put 31 Dec markers at the rail's 100% point, half outside the timeline against the card border, and the 20px circle left a 16px content box for a 17px label. Measured at 1440px and 375px.
2026-10-03Cadence Meter month labels switched from text.disabled (black at 38% alpha, 2.68:1) to the opaque text.secondary token (6.32:1 on the card). All twelve labels now render in one tint.#12738Bug #12738 — the labels failed WCAG AA 4.5:1. The Mar/Jul tint was not a designed state: every label had the same computed colour, and the translucent text anti-aliased differently at some sub-pixel offsets.
2026-10-02SSA (DOPS) WBA now shows the Procedure/Surgery lookup on create, matching PBA — WBA_PROCEDURE_ASSESSMENT and WBA_TYPE_CATEGORY_MAP both now classify SSA as a procedure assessment alongside PBA.#12701Live Task 9 audit (#11194) found the SSA Add-WBA form missing the Procedure field the spec's shared-registrar-section table requires for "PBA + SSA only", contradicting the DOPS=SSA "same form, dual label" directive (#5718).
2026-10-02CBD assessor review form no longer renders the CEX-only Focus of Encounter / Complexity of Case fields.#12702Live Task 9 audit (#11194) found FormCBD.ReviewSection.tsx was a near-verbatim copy of FormCEX.ReviewSection.tsx with the CEX-only fields left in; the spec's type-specific-consultant-fields table lists CBD's "Extra field" as none.
2026-09-29activities.getByType is now scoped server-side to the caller's own rows via enforceUserScope (super-admin stays unscoped). Previously it applied no user scope, so any signed-in user received every user's activities of the requested type — e.g. every doctor's WBAs and logbook cases — with patient identifiers decrypted at the read boundary. Added Security AC and @regression @issue-12585 BDD scenario.#12609, #12585Security — PHI exposure found by the adversarial review of PR #12570. Owner-only scope matches getById/getAll/searchSurgicalEntriesByUrn; the procedure has no frontend or mobile caller, so narrowing it breaks no UI flow.
2026-09-28activities.getById now permits a read by the WBA's ASSIGNED assessor — activities.assessorId equal to the caller's own server-resolved user id, for type = 'wba' rows only — in addition to the owner and super-admins — but only once the WBA has been SHARED (status pending / verified / accepted): an unshared draft stays owner-only even though the form already names its assessor. assessorId can only be changed by the owner (#12568), so a stranger can no longer make themselves the assessor to unlock this read. Every other activity type (reflection, feedback, audit event, logbook case) stays owner-only, and any other consultant or trainee is still FORBIDDEN. Fixes the 'Record not found' an assessor hit when opening a shared WBA from its notification or 'Sign off … WBA' to-do. The same shared-only rule applies to the list endpoints (getByFilter in every branch, getAll): a WBA owned by someone else is listed only once shared, whatever status filter the client sends (wbaVisibility.ts). activities.getByType is scoped to the caller by the separate #12585 hotfix (#12609).#12570, #12543, #12568A WBA is a two-party record (trainee + the consultant it was shared with) but getById enforced owner-only scope, so the assessor's deep-link read returned 403 and the whole review / sign-off flow was blocked (found in the Task 16 Notifications walkthrough, #11218). Security — the widening is matched on the stored assessorId against the session identity, never a client-supplied id. Covered by wbaAssessorRead.integration.test.ts (owner ok, assigned assessor ok, other consultant FORBIDDEN, other trainee FORBIDDEN, assessor of an unshared draft FORBIDDEN, non-WBA type FORBIDDEN for its assessor, unshared draft absent from getByFilter / userScope / getAll) and @issue-12543 BDD scenarios.
2026-09-28Received WBA(s) tab: (1) the Sender cell (and the Submitted tab's Assessor cell) now resolves the person's name from the getUsersForPicker directory when the owner-scoped getUserProfile read is refused, instead of rendering '—'; (2) each received row carries an always-visible, labelled button — 'Review' while the WBA is pending, 'View' once actioned — replacing the hover-only icon. Backend: sign-off is a dedicated server-side transition, activities.signOffWba (#12568) — callable only by the WBA's assigned assessor on a pending WBA — which writes the status, a competency_signoff audit row and the clear of the assessor's wba_submitted / msf_submitted to-do in ONE transaction (if the audit insert fails the sign-off is rolled back), then creates the trainee's 'has reviewed your WBA / MSF' notification server-side. Deleting a WBA now also deletes its own wba_ / msf_ submitted / reviewed notifications, so they no longer deep-link to a missing record. The mobile Received list (no sender, unlabelled action) is tracked in #12582.#12570, #12544, #12568Found in the Task 16 Notifications walkthrough (#11218): Sender showed '—' because getUserProfile is owner-scoped and the assessor is not the trainee; the Actions column looked empty because HoverActionsCell renders at opacity 0 until the row is hovered; no sign-off audit row was written anywhere (no caller of auditTrail.logCompetencySignoff existed); the to-do never cleared because the reviewed-status check only matched 'approved'/'rejected' while the WBA consultant forms write 'verified'; and the trainee's 'reviewed' notification was created only by the client, so an API sign-off never notified them. The first cut audited the sign-off inside the generic activities.update best-effort (skipped on draft→verified, swallowed on write failure) and let any writer's 'verified' clear the to-do; the #12568 guard plus the transactional signOffWba path replace that. The picker directory is already readable by every authenticated user, so the Sender fallback exposes no additional data.
2026-09-28activities.update now enforces WBA ownership: only the owner (or a super-admin) may update a type = 'wba' row (incl. MSF), the existing #5054 guard still blocks the owner once it has left draft, the status can only be written as draft / pending, and on EVERY activity type a non-owner can no longer change userId, assessorId, cpdSupervisorId or type. A new activities.signOffWba procedure is the only way a WBA reaches verified / accepted: it requires the caller to be the row's assigned assessor (not its owner) and the WBA to be pending, accepts only the review fields (attributes, conRatingLevel, feedbackConsultant), and commits the status, the competency_signoff audit row and the assessor to-do clear in one transaction. The WBA page, MSF page and Work Queue WBA / MSF review dialogs now call it. Adversarial-review follow-up (same PR): activities.create only creates a WBA as draft / pending (no self sign-off at birth); a record's type can never be changed into or out of wba through update (closes the draft wba → logbookCase → verified → wba hop), and any other type change is super-admin only.#12570, #12568Security (#12568): any authenticated user could set another user's WBA assessorId to themselves — which then unlocked the #12543 assessor read (cross-user read of the trainee's assessment) and the #12544 sign-off audit (a forged competency_signoff row) — or set its status to verified and silently clear the real assessor's to-do. Red on the pre-fix code and on origin/uat: wbaUpdateGuard.integration.test.ts (stranger assessorId tamper, two-call read escalation, stranger verify, stranger trainee edit, assessor verify via update, owner self sign-off of a draft, non-WBA re-pointed as a WBA) and wbaSignoffAudit.integration.test.ts; @issue-12568 BDD scenarios. The review follow-up adds red→green + mutation evidence for create-as-verified, the type hop, and an owner who names themselves as assessor (the self-sign-off test now reaches the self-sign-off clause).
2026-07-17Close Q21 (Reflections model): Matt D8178.8 (2026-07-13) confirmed Reflections are free-text per the original 'PPP - Retrospective Evaluations' spreadsheet (Reflection Tool tab: Case Summary / Situation / Learning, each a free-text field), and the Pendleton model belongs to the Feedback tool only. Flipped the status-table 'Reflections — Pendleton model' row from ❓ OQ to ✅ Resolved. No form change: the shipped ReflectionForm already matches the spreadsheet's free-text structure.#10759, #9653D8178.8 answered — the spec's engineering default (Pendleton = Feedback only) is now Matt-confirmed rather than open. Drift-audit compare against the source spreadsheet found the current form already correct; Matt asked to agenda a call to run through the current build (run-through sign-off remains open, issue keeps for-discussion).
2026-07-12Mark 360 min-reviewer enforcement (PPP_360_MIN_REVIEWERS=7) as shipped; inputSchema.ts superRefine confirms server-side enforcement. Note role composition check (Professional Supervisor + Colleague Appraiser) remains an open question — count validated, roles not.#10513, #10484, #5009C13 drift: spec showed ⚠️ Bug for #5009 but reviewer-count enforcement was already shipped server-side.
2026-07-11Confirmed/cited MCR/MSF/360 participant model against shipped code (closed Q23); raised the "Named Professional Supervisor" rename as open Q22 (do-not-infer, awaiting Matt).#10368, #8171Matt confirmed the MCR/MSF/360 participant model (#8171, 2026-06-30, restated 2026-07-10). A #10368 code audit verified all three match the shipped implementation (mcrService + FormMCR, FormMSF, ppp360 router + verify360ReviewerToken). The one genuinely-unresolved item — whether "Named Professional Supervisor" is a display-rename of the existing Educational Supervisor role (QMF→MSF precedent, enum/FK unchanged) or a new distinct role — stays open pending Matt, per the appraisals-wba "do not infer" rule.
2026-07-07Reflections §: appraisals.reflections.tag-to-logbook-entry flipped ⏳ → ✅. A standalone reflection can be tagged to a specific surgical logbook entry; the linked case is now echoed back as an explicit read-only "Linked to: …" confirmation. Added executable @smoke @issue-10134 BDD scenario.#10139, #10134, #6399Split from #9653 (the additive, non-blocked half). The tag picker itself shipped earlier via #6399 (LogbookTagField → attributes.logbookId, with delete-time link scrubbing #5051), so the AC's write path was already satisfied — this PR closes the governance gap (AC left ⏳) and adds the missing read-side traceability: the tagged case is surfaced back to the user as a confirmation, plus a deterministic component test and an executable smoke scenario. The Pendleton-model field structure stays out of scope in #9653 (OQ Q21 / D8178.8, pending Matt).
2026-06-30Corrected Pendleton model attribution — Pendleton is Feedback only, not Reflections; removed incorrect AC from Reflections sub-feature, fixed misattributed source citation on tagging AC, added OQ Q21, and marked BDD scenario as @wip pending Matt confirmation.#9852, #9827The prior AC "Reflections follow Pendleton feedback model" was inferred from transcript-2026-05-19 §PendletonModel (~00:43:00) — "platform supports the Pendleton model of feedback to create safe, implementable feedback loops" — but investigation (#9827) confirmed this quote is in the Feedback tool context, not Reflections. Engineering default is Pendleton applies to Feedback only; Matt must confirm before Pendleton is implemented in Reflections. The "tag to log entry" AC source was also corrected from §PendletonModel to §19 (the two requirements were erroneously sharing the same incorrect citation).
2026-06-30Flip Q20 (MCR/360 collaboration window) from Open → Answered. Default = 7 business days (shipped in #9234 via migration 0178), kept super-admin configurable (#2887). Records the 2026-06-30 confirmation and removes the stale "pending Matt" / 48h framing.#9807, #9059Matt approved extending the window (#8170) and delegated the exact value to config; the team confirmed 7 business days on 2026-06-30. Business days (not calendar) is the settled default, preserving the Friday→Monday weekend-coverage intent that motivated Q20. Additive and fully reversible via the super-admin setting.
2026-06-29Appraisals-wba data-sharing — backend persistence + appraiser-read enforcement (follow-up to #9649). New appraisals.sharing_preferences jsonb column (migration 0182) + appraisalSharingService.persistSharingPreferences (owner/draft-only, wired into Save Draft / Submit) + getSharingPreferences. getAppraisalSummary now filters the appraiser-only view server-side via filterAppraisalForAppraiser — withhold-by-default, the doctor (owner) always sees everything, identity from ctx (no IDOR). Flips appraisals.yearly.data-sharing-opt-in persistence + enforcement to BUILT.#9783, #9764#9649/#9762 shipped the doctor-facing opt-in control + the filterSharedSections seam but deferred backend persistence and the appraiser-read enforcement to #9764. This lands both so the doctor's per-element opt-ins take effect end-to-end — the appraiser only ever sees the elements the doctor opted to share (decision D8178.6, Matt "Option A"). New audit action appraisal_save_sharing_preferences is a varchar value (no enum migration).
2026-06-29Add the appraisal "runner" completion-nudge (appraisals-wba.runner.completion-nudge) — reminds doctors to finish their annual appraisal 3 / 2 / 1 months before the completion-due date, with the schedule DB-backed (appraisal_reminder_settings) and the Dec–Jan window as the configurable seeded default.#9768, #9650Epic #9647 child #9650. Matt 2026-06-24 — keep the Dec–Jan window as the default but do not hardcode it so regional variance can be configured; relative-to-due 3/2/1-month model per Matt's 2026-06-28 roadmap edit.
2026-06-29Add user-controlled data-privacy sharing to the yearly appraisal — a "Data Sharing" step with per-element opt-in switches (Personal Details, Year in Review, Well-being, CPD, MSF/360 & competency, PDP), each OFF by default. New utils/sharing.ts (getDefaultSharingPreferences/filterSharedSections) plus the sharing field on appraisalFormSchema enforce withhold-by-default. AC appraisals.yearly.data-sharing-opt-in ⏳ → ✅.#9762, #9649Decision D8178.6 — Matt confirmed "Option A" (#9649, 2026-06-29): sharing defaults to WITHHOLD / opt-in (private by default). Safe default for clinical PII — opt-in cannot accidentally over-share with the appraiser; the doctor explicitly opts each element in.
2026-06-29Convert the annual Appraisal Form's free-jump tabbed/stepper layout into a guided progress-indicator modal flow — a determinate completion meter ("N of M sections complete") plus per-step completion ticks that advance as the doctor works through the sections. New AC appraisal.form.progress-indicator-modal-flow.#9761, #9654, #3956Yearly Appraisal UX epic #9647 / design intent #3956 (transcript 2026-03-25 §AppraisalModal — "tabbed interface replaced with modal dialogue box that shows indicators of progress").
2026-06-29Rename "Educational Supervisor" → "Professional Supervisor" in all AC descriptions for non-consultant appraisal flows.#9718, #9703Vernacular change requested by Matt (#9703) — "Professional Supervisor" is the correct term for non-consultant profiles platform-wide.
2026-06-24Widen the default MCR/360 collaboration window from 5 → 7 business days (systemSettings.collaborationWindowDays, migration 0178). The window stays super-admin configurable (#2887); mcrService.createMCR / extendCollaborationWindow fallbacks move 5 → 7 to agree with the new DB default. Q20 reframed (configurable business days, not a hardcoded 48h) and recorded as OPEN.#9234, #9059Reports started on a Friday risked falling outside the window before the consultant body saw them on Monday. Matt approved extending the window (#8170) but has not confirmed the exact value — 7 business days is an engineering DEFAULT (additive, reversible via config) pending Matt's sign-off on the value and the calendar-vs-business-days unit; spec Q20 stays open and the issue/PR carry for-discussion.
2026-06-24Add the deferred foreign-key constraint on appraisal_pdp_items.source_smart_goal_id → ppp_smart_goals.id (migration 0177, ON DELETE SET NULL). Restores the ORM-level .references() spun out of #9167 and adds the matching Postgres constraint with a DO-block validator that fails the deploy if the FK is absent post-migrate.#9220, #9175Without the FK, deleting a Roadmap SMART goal left orphaned source_smart_goal_id values pointing at non-existent rows. ON DELETE SET NULL nulls the referencing PDP rows instead, hardening referential integrity behind appraisals.yearly.roadmap-crossover.import-active-goals (#9095) while keeping the signed-off appraisal archive immutable (no cascading deletes). No user-facing behaviour change.
2026-06-23Appraisals-wba S4 — Roadmap ↔ appraisal SMART-goal crossover ("linked, not merged", #8167 Q10). New appraisalRoadmapCrossoverService — importActiveRoadmapGoals (owner+draft guard, dedupe on sourceSmartGoalId; SMART fields → PDP narrative columns; wired into rollIntoAppraisal at review time + the "Import from Roadmap" PDP action), endorsePdpGoal (appraiser+under_review guard sets appraiserEndorsed), and exportEndorsedGoalsToRoadmap (called inside completeAppraisal after the sign-off transaction; endorsed + not-yet-on-Roadmap goals become new active SMART goals via createSmartGoal; failures swallowed so a Roadmap-sync hiccup never blocks sign-off). appraisal_pdp_items gains source_smart_goal_id + appraiser_endorsed (migration 0176). Flips appraisals.yearly.roadmap-crossover.import-active-goals and .export-endorsed-goals to ✅.#9167, #9095Resolves ppp-roadmap Q10 / #8167 ("linked, not merged") and the #8178 crossover spec — SMART goals live on the Roadmap year-round; the yearly appraisal pulls active goals in at review time as a point-in-time PDP snapshot and pushes appraiser-endorsed new goals back on sign-off, keeping the signed appraisal a legally-stable locked archive. Builds on the S3 (#9094) PDP/SMART-goal surface. Migration 0176 adds the two crossover columns (+ .validate.sql); two new audit actions are varchar(100) (no enum migration).
2026-06-23Appraisals-wba S3 — appraisal record + PDP/SMART-goal persistence + CPD-review roll-in + appraiser discussion. New appraisalPdpService (savePdpItems owner/draft-only set-replace, getPdpItems for doctor+appraiser, addPdpDiscussionComment on the pdp section); rollIntoAppraisal procedure wires the S1 rollAnnualReviewIntoAppraisal so completing the CPD-Home annual review creates the yearly appraisal + PDP and links them; useAppraisalData now persists pdp.goals on save/submit (previously discarded). Flips appraisals.yearly.review-rolls-into-appraisal, logs-as-two-separate-things, and pdp-discussed-with-appraiser to ✅.#9164, #9094Matt's CPD-Home → Appraisal → PDP crossover (Dev Session 2026-06-03, #8178). The appraisalPdpItems table shipped in #9092 but nothing persisted to it; this is the user-facing PDP/SMART-goal creation + appraiser-discussion slice the annual review rolls into. Roadmap import/export is out of scope (#9095/S4, held). No migration (table pre-existing; new audit actions are varchar(100)).
2026-06-23CPD-Home annual CPD-plan review compliance step (appraisals-wba S2) — RequirementTile + review dialog drive the S1 state machine not_started → in_progress → completed; getAnnualCpdReview / startAnnualCpdReview / completeAnnualCpdReview procedures (self-scoped, audited). Flips appraisals.yearly.annual-cpd-review-compliance-step to ✅.#9151, #9093Matt's "two separate things, linked" model (Dev Session 2026-06-03, #8178) — the annual CPD-plan review is a CPD-Home compliance step logged distinctly from the yearly appraisal it later rolls into (#9094). Builds on the S1 data model + state machine (#9092).
2026-06-22Yearly Appraisal annual-CPD-review data model + state machine (build slice S1) — appraisal_annual_cpd_reviews table, annual_cpd_review_status_type enum, appraisals.annual_cpd_review_id linkage column, the forward-only state machine, and the appraisalWorkflowService shell.#9127, #9092, #8178, #9091First build slice off spec enabler #9091, implementing the foundational persistence + transition layer for Matt's "two separate things, linked (not merged)" model (#8178, transcript 2026-06-03). The annual CPD-plan review log is a distinct record from the appraisal it rolls into; the new linkage column records the relationship without merging the rows. Unblocks #9093 (CPD-Home review step), #9094 (appraisal+PDP), #9095 (Roadmap crossover).
2026-06-22Author the Yearly Appraisal annual-CPD-review × appraisal × PDP crossover spec section + AC scaffold (spec enabler); add @smoke + @manual BDD scenarios.#9091, #8178, #8167Spec-first enabler (#9091) for the not-yet-built Yearly Appraisal (#8178). Documents the CPD-Home annual-review→appraisal→PDP roll-in ("logs as two separate things") and the Roadmap "linked, not merged" SMART-goal import/export (ppp-roadmap Q10 / #8167). Every AC is Matt-sourced (transcript 2026-06-03) or the resolved Q10 design decision — no inferred requirements. Unblocks build slices #9092–#9095.
2026-06-20Default WBA Performance Overview widget to open on first visit; persist collapse choice per user in localStorage.#8991, #8949UAT: mobile user (iPhone) found the widget collapsed by default with no indication content existed — #8949. WidgetSection now defaults to expanded on all viewports and remembers the user's explicit collapse/expand choice via localStorage keyed by sectionId + userId.
2026-06-10Added code-paths for wba.ts and wba-role-config.ts#8606, #8485wba-role-config.ts is already cited in appraisals-wba.md as single source of truth (#6359); drift gate now fires correctly when WBA constants change, per #8485
2026-06-09Show the "Collated by representative" disclaimer notice on the MCR consultant form (it was on MSF only). Notice wording extracted into a shared builder consumed by both FormMSF and FormMCR; MCR Submit gates on the acknowledgement.#8346, #8174Spec↔code drift from the SGD audit (appraisals Q15) — the spec says both MSF/MCR forms must show the notice but it was missing from MCR.
2026-06-03Resolve Q14/Q15/Q16 to Closed Questions (CEX focus-of-encounter constant, MSF header/texts, representative notice) — code-confirmed; MCR-notice gap tracked as #8174.#8166, #8174SGD governance sweep — transcript-grounded audit found these questions already answered; reconciled the spec to shipped reality and moved them out of Open Questions.
2026-06-02WBA appraisal reminders (MSF/MCR/360) now distribute across the doctor's selected cadence period (cadencePeriodMonths, default 6) instead of a fixed 12-month plan. calculateAppraisalDueDates's backward branch no longer anchors to a hardcoded month 12, and processAppraisalNotifications threads cadence_period_months through. Also corrects the wba.cadence.set-during-onboarding AC to describe the two engines accurately (appraisal subset → reminders; WBA subset → roadmap milestone-engine).#8138, #8118processAppraisalNotifications ignored cadencePeriodMonths, so period-6 doctors were reminded at half their intended appraisal cadence. First-period-only distribution chosen by Rob (acting PO, 2026-06-02). The AC correction restores the #8119 review fix lost when auto-merge raced a late commit.
2026-06-02Flips wba.cadence.set-during-onboarding ⏳ → ✅. The WBA/appraisal cadence (CBD, CEX, PBA, MCR, 360, MSF) is collected in the unified Cadence Settings card via the CadenceTracker (selectable 3/6/12-month period) alongside the surgical targets, and those values feed processAppraisalNotifications, which reads cadenceTracker to compute appraisal due dates.#8119, #8041Verify-and-close: already shipped. The #8041 deep-dive assumed an orphaned CadenceSettingsStep needing re-wiring, but the WBA CadenceTracker is already live in the My Profile cadence card and wired to the reminder engine. Residual: the engine ignores cadencePeriodMonths (treats per-period counts as annual) — split to #8118.
2026-06-02Reviewer-non-response flow — two automated reminders (~1wk, ~2wk) to a non-responding supervisor/appraiser, then a requester "prompt again" control after ~3 weeks that sends a friendly-reminder notification only (no extra action item) and notes the reviewer has already had two automated reminders.#8095, #7916Matt (2 Jun 2026, #7916) decided PPP nudges but the doctor is ultimately responsible (Q1=C), a 1-week + 2-week automated cadence plus a requester-driven prompt (Q2=B), and no "PPP is not responsible" disclaimer (Q3=No). Built on the existing appraisalReminderService + notifications path; cadence thresholds are configurable constants defaulting to 7/14/21 days.
2026-06-01WBA/appraisal cadence is collected during onboarding as part of the unified Cadence Settings step (wba.cadence.set-during-onboarding), not surgical targets alone; these values feed the appraisal/WBA reminder engine.#8005, #7344Matt (#7344) — "Our cadence settings includes WBA's and Appraisals." The richer assessment-cadence step already exists (CadenceSettingsStep) but was unplugged from the wizard; the agreed model (Rob acting PO, 2026-06-01) re-wires it alongside surgical targets.
2026-05-28WBA page Received Appraisals widget now uses the canonical thin-doughnut style (inner 75 / outer 120 desktop, 55 / 90 mobile) instead of the previous thicker ring (60/100 desktop, 70/120 mobile). The radii are exported as named constants and guarded by a unit test.#7831Matt reported on ppp-dev that the doughnut on /wba?tab=1 was visibly thicker than the doughnuts on every other widget (CPD, Surgical Logbook, Appraisals, Timesheets, Retrospective, sibling WBA Actioned). The WBA chart now matches the rest of PPP for visual consistency. No behavioural / data change.
2026-05-28WBA filter bar now uses the shared pill-based date-range filter (PillDateRangeFilter) so the WBA page matches the visual treatment of the PPP Pulse chart range switcher on the dashboard. Presets exposed are 3 Months / 6 Months / 1 Year / All Time / Custom; selecting Custom opens a popover with start and end date pickers (previously the popover was reached via a separate "Select Date Range" button, with chip-style presets next to it).#7813UI consistency. The existing Button + Chip combination diverged visually from the Pulse pill switcher, even though both surfaces drive a date window. The shared component lives at instances/ppp/src/components/PillDateRangeFilter/ and reuses the same pill toggle sx tokens as the Pulse chart (pillToggleGroupSx), so future date-range filters can adopt the same look without duplicating styles.
2026-05-28Filter bars now consume the shared PillDateRangeFilter; doughnut widgets share the THIN_DOUGHNUT_CONFIG from the canonical ThinDoughnutChart component.DRY refactor — eliminate duplicated chip-preset and chart-radius logic so a single source of truth governs the date-range pill and thin-doughnut look across the app.
2026-05-27Condensed the WBA create/edit modal to brand density (design-token spacing, gaps reduced from 30px to SPACING.base/sm) so the form is fast to scan and fill, and replaced the confusing three-button close confirmation (Cancel/No/Yes) with two clear choices — "Discard changes" (destructive) and "Keep editing" (primary).#7817Rob flagged during the 2026-05-27 demo session that the WBA create modal had too much whitespace and was slow to fill, and that the unsaved-changes close dialog showed three buttons where "Cancel" and "No" read as duplicates. The shared UnsavedChangesDialog gained an opt-in twoButton variant (default off) so only WBA adopts the two-button form — the ~12 other consumers keep the existing three-button behaviour and are unaffected.
2026-05-27Four confirmed WBA/appraisal form sign-offs — CEX rating items 6/7 reverted to source wording (#6995); MSF closing sign-off fields added (specific examples, probity &amp; health concerns, environments observed) + display rename QMF → MSF (#6999); PBA/SSA global rating confirmed on the existing 9-level a/b scale, now guarded by a regression test (#7001); MCR keeps all 9 capabilities with a single off-switch for AU-optional items 8 &amp; 9 (#6998).#7727, #6995, #6999, #7001, #6998Matt confirmed each change against the WBA Forms spreadsheet during the 2026-05-27 sign-off pass. CEX q6/q7 had drifted from source and near-duplicated q4; MSF was missing its three closing fields; PBA/SSA already rendered the confirmed 9-level a/b picker (the #5720 audit premise that it used the 5-level scale was a false alarm — FormSSAPBA always consumed the 9-level DOCTOR_RATINGS from constants/doctor-ratings), so #7001 adds only a regression test; MCR items 8 &amp; 9 are GMC/UK extensions that AU enterprises may opt out of, so they are gated behind a flag rather than removed. Persisted enum/identifier values (wbaType 'qmf', QMF_* code names) were intentionally left unchanged to avoid data migration risk.
2026-05-20Cadence Meter now distributes a non-consultant's required WBAs across the active cadence period instead of bunching them on a single window-close date. generateCadenceMilestones emits one checkpoint milestone per required assessment (e.g. CEX target = 3 → 3 evenly-spaced checkpoints), and short-circuits for users with userType === 'consultant' so the cadence meter no longer produces milestones for fully-qualified consultants.#TBD, #6356Closes wba.cadence.meter-distributes-roadmap (#6356). The previous engine collapsed every WBA type to a single end-of-period milestone, which did not satisfy the "distributes required assessments across the roadmap" acceptance criterion from Dev Session 2026-04-23 §CadenceMeter, and ran the cadence calculation for consultants who have no trainee cadence.
2026-05-19WBA feedback review now presents three sequential boxes (Feedback → Ratings → Global Summary) with progressive disclosure: the Ratings box stays locked until a Feedback question is answered and the Global Summary box stays locked until a Ratings question is answered.#6555, #6355Implements AC wba.form.three-sequential-boxes from the 2026-05-19 Dev Session transcript (§WBAFeedbackForm ~00:58:37 — "three distinct boxes; next section only activated after user answers question from previous section"). Locking is shared across all WBA types (CEX, CBD, SSA/PBA) via SharedReviewSection so the stepwise contract cannot drift per type.
2026-05-19Add role-based WBA configuration grid (WBA_ROLE_CONFIG_GRID) defining which forms, statuses and critical conditions apply to consultants vs non-consultants — single source of truth in packages-shared/ppp, re-exported through the backend WBA dashboard config.#6554, #6359Role gating for WBA forms/statuses/critical conditions was implicit and scattered. Matt specified a grid distinguishing consultants from non-consultants (transcript 2026-05-19 §RoleConfigTable ~00:40:47).
2026-05-19Competency and critical-condition widgets pre-populate all items (every competency level / specialty critical condition) before any data entry, rendered as horizontal progress bars instead of donut charts, with a dedicated level-4 progress bar tracking the 3 required level-4 assessments.#6553, #6358AC wba.competency.preloaded-horizontal-bars (transcript 2026-05-19 §CompetencyDisplay) was not built — the widgets used donut charts that only rendered non-zero data, so items were invisible before data entry. Pre-populating all items as horizontal bars and surfacing the 3 × level-4 requirement closes the gap.
2026-05-19Implement the appraisal evidence pack PDF export. New tRPC mutation appraisalWorkflow.exportEvidencePack aggregates personal details, CPD, MSF/360, WBA, mandatory training, PDP, appraiser comments and outcome into a single server-side @pdfme/generator PDF; the /appraisals/evidence-pack route now renders an export panel instead of the coming-soon placeholder.#6542, #5011, #5096GMC revalidation and AHPRA CPD audit both require a submittable consolidated evidence document; no appraisal-to-PDF service existed (#5011), and the route previously showed only a placeholder (#5096).
2026-05-19WBA summary widget now distinguishes submitted-but-not-complete WBAs from completed ones: completed (verified) segments render with a solid fill, submitted-not-complete segments render with a 45° hatch pattern, plus an in-UI pattern legend. getWBASummaryData breakdown gained a submitted count.#6538, #6357Implements AC wba.widget.hash-colour-submitted from the 2026-05-19 Dev Session transcript (§WBAWidgetColours ~00:32:06 — "hash colors for submitted but uncompleted; full colors for completed items"). Pattern + colour encoding satisfies WCAG 2.1 AA so the distinction is perceivable without relying on colour alone.
2026-05-19WBA created from a logbook entry now auto-populates case security (case acuity) and principal procedure from the linked entry. AC wba.form.auto-populate-from-logbook flipped to done.#6504, #6353Implements transcript requirement (Dev Session 2026-03-25 §WBADataPull ~00:26:17 — "case security and principal procedure pulled from log book entry to WBA form"). createWbaFromLogbook previously copied urn/date/ hospital/assessor/user only; caseAcuity and procedureId are now carried so the WBA form pre-fills these fields while leaving the rest user-editable.
2026-05-19Verified wba.reflection.no-duplicate — a logbook event that already has a reflection no longer offers "Complete Reflection" again; AC flipped to ✅ with @regression coverage.#6497, #6354Issue #6354 — system must prevent adding a second reflection to a logbook event that already has one (transcript 2026-03-25 §ReflectionDedup: "reflection not available if already selected for that log book event"). Spec-drift audit found this already implemented: both Consultant and NonConsultant CheckboxOptions.tsx replace the "Complete Reflection" checkbox with a link to the existing reflection when a successful related reflection form exists for the logbook event. Added a @regression @issue-6354 scenario to lock the behaviour and flipped the AC to ✅.
2026-05-19Enforce minimum 7 reviewers on 360 review submission — draft saves exempt; Zod schema + service guard + frontend validation.#6301, #5009Single-reviewer 360 reviews violated statistical validity requirements for multi-source feedback. Matt confirmed minimum of 7 raters on 3 May.
2026-05-19Drift re-baseline — BDD lifecycle scenarios added for 3-phase yearly appraisal; spec status flipped to specified (Q1–Q9 answered, Q6 resolved by code audit).#6295, #5341All open drift questions from the original #5341 scope answered in PR #5644. Adding smoke and regression BDD scenarios for the confirmed Draft → Review → Signed-off lifecycle closes the last remaining gap before the status flip.
2026-05-19SGD transcript audit — 13 new ACs added from 2026-05-19 Dev Session transcript; 3 new Open Questions (Q17–Q19); notion-last-synced updated to 2026-05-19.#~Transcript-to-spec audit following the SGD protocol (docs/features/_sgd-audit-protocol.md). Source: raw Gemini notes for Catchup 2026-05-19 06:30 BST. Every AC is cited with a paragraph-level <abbr> hover reference to the transcript passage it was derived from. No assumptions made — ambiguous passages raised as Q17–Q19 rather than resolved silently.
2026-05-06WBA drift audit — side-by-side comparison of code fallbacks vs spec table vs Matt's spreadsheet; 7 confirm-with-matt deltas documented as Q10–Q16.#5791, #5720Matt re-shared WBA Forms spreadsheet on 5 May 2026. Audit reveals code fallbacks diverge from the spreadsheet reference in several places (CEX q6/q7 labels, MCR Part 1 count, QMF end-section fields, PBA/SSA global-summary sub-levels). Per issue scope, no code changed — deltas are raised as Matt-confirmation questions only.
2026-05-05Apply Matt's 3 May Notion answers to ACs that contradicted them — rater identities NOT anonymous (#5008 flipped, closed); 360 minimum 7 + ES + CA roles (#5009); DOPS = SSA UI alias (replaces closed #5024); appraiser dropdown filter is appraisee-type-aware (#5063 superseded — ES/CA synonymous).#5713, #5008, #5009, #5024, #5063Earlier ACs were locked when engineering proposed anonymisation, single-reviewer-360 enforcement, and Consultant-only filter; Matt's 3 May Notion replies flipped or refined each.
2026-05-05Reframe Q4 — WBA forms are already built; spreadsheet is an audit reference (not requirements); Moreton Blue UAT iteration may have moved things without spreadsheet update; #5720 rewritten as drift audit not from-scratch build.#5713, #5720Original framing assumed engineering needed to data-model from scratch. In fact forms/WBAForm, forms/WBAConsultantForm/{FormCBD,FormCEX,...}, MCRForm, MSFForm all exist; prompts are DB-driven via Settings → Form Prompts (#2283). Rob clarified 5 May that Matt iterated during Moreton Blue without documenting, so spreadsheet must not be auto-applied.
2026-05-05Q6 (Evidence-pack PDF priority) resolved by code audit, not engineering ordering. #5090 closed as already-fixed (SVG bug fixed by #5229/#5271); #5096 rescoped to small placeholder route; #5011 description updated — not from scratch, reuses existing pdfme/triggerDownload pattern.#5713, #5090, #5096, #5011Original engineering-led answer assumed all three issues were live. Code audit on 5 May showed two of three were stale. Closing/rescoping prevents wasted work and clarifies the only real implementation issue is #5011.
2026-05-05Add WBA Form Data Model section sourced from Matt's WBA Forms spreadsheet — schemas for CEX, CBD, PBA, SSA, Quarterly MSF, MCR. Implementation tracked in #5720.#5713, #5720Q4 (Matt 3 May) asked for the spreadsheets to be shared so engineering could data-model. Spreadsheet shared 5 May; data-modelled directly rather than waiting for a live walkthrough.
2026-05-05Q6 (evidence-pack PDF priority) resolved engineering-led — sequence is #5096 routing fix → #5011 export endpoint → #5090 SVG bug. Natural dependency chain.#5096, #5011, #5090Matt asked "what's this one?" on 3 May; engineering can sequence the dependency chain without further input.
2026-03-11Transcript audit 2026-03-11 — §37 classified F (feedback); Q20 added to Open Questions about MCR/360 collaboration window duration.#~§37 of the 2026-03-11 Dev Session transcript ("Platform overview — first full review") surfaced feedback that the 48-hour MCR/360 collaboration window is too restrictive for weekend submissions (reports started on Friday are outside the window by Monday). This is an unresolved policy question — no AC was changed. Added as Q20 to track the decision on window duration and push notification approach. All other paragraphs in the transcript that touched appraisals were classified C or CONFIRM.
2026-04-21Triage sweep — 21 open issues mapped to this spec#5262, #5096, #5095, #5094, #5093, #5092, #5091, #5090, #5088, #5063, #5054, #5026, #5025, #5024, #5023, #5022, #5011, #5010, #5009, #5008, #4977Unblocks drift detection on the features index.
Jan 2026WBA chart legend (required vs completed) identified as gapWalk-through notesCharts exist but missing context
Jan 2026Consultant in-app guidance identifiedWalk-through notesConsultants unsure what to do on review screens
Jan 2026MSF / 360 blank dialog bugs fixed#3579, #3580Notification deep-links not implemented
Apr 2026WBA add form now blocks Save/Submit when required fields (WBA Type, Date, Hospital, Assessor, Setting) are empty — prevents blank draft records#5021Submit path bypassed RHF validation so empty records were being persisted
Apr 2026Appraisal Save Draft and Submit now call the real tRPC mutations (appraisalWorkflow.createDraft and appraisalWorkflow.submit) instead of faking success — data is no longer silently lost#5007Handlers in useAppraisalData.ts were stubs that displayed a success snackbar without hitting the backend
Apr 2026MSF request form now blocks Save/Submit when Hospital, Assessor, or Job Duration are empty — prevents orphaned MSF records#5089Form previously accepted fully empty submissions
2026-04-21activities.getByFilter now enforces server-side scope on the assessorId filter — non-super-admin callers must set assessorId to their own user id or receive FORBIDDEN. The WBA "Received" tab gate in WBA/index.tsx was cosmetic only; any authenticated user could previously list WBAs addressed to another consultant by calling the tRPC procedure directly.#5027Security — clinical data confidentiality. Frontend role gates must be mirrored server-side (see auth-zero-trust-routing.md).
2026-04-22Drift re-baseline — 20 open issues absorbed into ACs, Current Status, Open Issues tables; ACs tightened to ≤15 words/bullet; Q5–Q9 added for Matt#5341, #5262, #5096, #5095, #5094, #5093, #5092, #5091, #5090, #5088, #5063, #5054, #5026, #5025, #5024, #5023, #5022, #5011, #5010, #5009, #5008Per /drift-fix appraisals-wba — see audit docs/features/.drift-audits/appraisals-wba.md. Status remains partial-drift pending Matt Q1–Q9.
Sources & Recordings ​
SourceNotes
https://drive.google.com/file/d/1HDF4acVA0QKImQ29Wu0rp_kVIgb-VGFVg_XwcZnUV6EDescription pending SGD audit
https://docs.google.com/document/d/1YhgQSFL9iOlQLoTrQYSwxRNXH4Dg4wefkItcwljQZmoDescription pending SGD audit

Internal documentation — org members only