Prompt 15 — Personal development plan, coaching, and adaptive follow-up

Required context

• Results experience
• PDP privacy/sharing model
• Adaptive recurrence override rules
• Current PDP screen and external development docs

Prompt

Build a development workflow that turns feedback evidence into two or three focused goals, actions, and follow-up questions without making the PDP an involuntary performance record.

1. PDP creation flow

From results, allow the subject to:

• Select or write a development theme
• Review supporting evidence and uncertainty
• Define desired observable behavior/outcome
• Create specific actions and support needs
• Set target/review date
• Choose private or shared visibility per field/goal according to policy
• Select coach/manager where allowed
• Confirm follow-up cadence and question linkage

Do not automatically create a goal from one low score. AI may suggest, but the subject or authorized owner must confirm.

2. Goal model

Support:

• Title and capability/construct link
• Outcome statement
• Actions/experiments
• Evidence criteria
• Support/owner
• Start, target, review dates
• Status: draft, active, paused, completed, archived
• Privacy/sharing state
• Linked question families/variants
• Recurrence override and follow-on cooldown
• Check-ins and activity history

3. PDP follow-up sequence

Rotate appropriate check-ins rather than repeating one rating:

• Commitment/action planned
• Action completed
• Learning/reflection
• Support/blocker
• Observer evidence
• Outcome
• Sustainability

Allow a weekly action check and a monthly/quarterly observer behavior check to coexist.

4. Recurrence overrides

Implement approved modes:

• Replace base cooldown
• Shorten only
• Lengthen only
• Suppress until milestone
• Force next once

force_next_once must:

1. Place the linked family in the next eligible survey only once.
2. Never duplicate it in the current instance.
3. Consume only when successfully presented.
4. Revert to follow-on cooldown.
5. Respect relationship applicability, observation minimums, privacy, and fatigue caps.

5. Development dashboard

Show:

• Active goals and next actions
• Progress/check-in timeline
• Evidence collected
• Upcoming feedback follow-up
• Shared participants
• Privacy status
• Blockers/support requests
• Completed/archived goals

Avoid gamified streaks that pressure disclosure. Use progress states that reflect actual actions and evidence.

6. Manager/coach collaboration

Only show explicitly shared goals/actions or policy-authorized content. Support comments, commitments, and check-ins without exposing private notes or full reports.

Allow the subject to review and change sharing where policy permits. Audit share/revoke events.

7. AI-assisted conversation

If released:

• Ground suggestions in visible result evidence and the shared glossary.
• State that suggestions are not HR decisions.
• Allow edit/reject and show source links.
• Do not send private content to an unapproved model/provider.
• Record model/version/policy metadata without logging sensitive prompt content where prohibited.
• Provide a non-AI path.

8. Tests

• Goal privacy/share/revoke
• Force-next-once and follow-on cooldown
• Pause/complete removes overrides
• Observation-window restrictions
• Mandatory overflow behavior
• Manager scope changes
• AI opt-out/failure
• Export/delete/retention policy

Acceptance gates

• A subject can move from evidence to a focused plan and next action.
• Private goals remain private in API, UI, logs, exports, and analytics.
• Recurrence behavior is deterministic and tested.
• The workflow supports development without automatic performance classification.