View on GitHub

campaign-mode

Work with a party of AI advisors with genuinely different perspectives.

ADR-CM-005: Campaign Mode Selection

Field Value
Decision ID ADR-CM-005
Initiative Campaign Mode
Proposed By Chris Barlow
Date 2026-02-14
Status Accepted

WH(Y) Decision Statement

In the context of Campaign Mode’s quest lifecycle and the range of users who may adopt it,

facing the reality that users approach campaigns with fundamentally different priorities – some want personal growth and self-discovery, some want to ship work efficiently, and some want both – and that the current design implicitly assumes a balanced orientation with no way for the user to express their intent,

we decided for three named campaign modes (Grow, Ship, Grow & Ship) selected by the user during Phase 1 (Quest Definition), with each mode tuning all NPC agent behaviour through mode-aware context,

and neglected a single fixed mode (forces all users into one framing), per-interaction mode switching (inconsistency mid-campaign), and a continuous slider (too abstract),

to achieve user agency over their campaign experience, appropriate NPC behaviour calibration, and alignment with SDT’s autonomy principle,

accepting that all three modes must be maintained in every NPC skill (increasing complexity), and that the mode distinction adds a decision point at campaign start.

Context

Campaign Mode currently treats all campaigns identically – an implicit assumption that every user wants the same balance of learning and productivity. In practice, users approach campaigns with fundamentally different priorities:

  1. Learning-oriented users want self-discovery, reflection, and personal transformation. The journey matters more than the destination. They want Gandalf to ask deep questions, the Guardian to assess understanding, and the Dragon to test growth.

  2. Productivity-oriented users want to ship work efficiently. Campaign Mode’s structured approach and diverse NPC perspectives help them cover blind spots and produce better deliverables. They want efficiency, not reflection.

  3. Balanced users want both – growth through doing real work. They want the full campaign experience without artificial separation of learning from delivery.

This aligns with Self Determination Theory’s autonomy principle: users should have agency over their own experience. Forcing a single orientation undermines the autonomy that makes Campaign Mode effective.

The User as Protagonist

This decision also addresses a gap in Campaign Mode’s framing: the documentation describes 10 AI agents but never explains what the human does. The user is the protagonist – the decision-maker who drives the quest, invokes agents, produces work, and faces NPCs. Mode selection is the first expression of this agency: the user chooses how they want to approach their campaign before the campaign begins.

Options Considered

Option 1: Three Named Modes – Grow, Ship, Grow & Ship (Selected)

Three distinct modes selected at campaign start, each tuning NPC behaviour through mode-aware context passed to all agents.

Pros:

Cons:

Option 2: Single Fixed Mode (Rejected)

All campaigns use the same orientation (implicitly Grow & Ship).

Why rejected: Forces all users into one framing. Productivity-oriented users find the reflection unnecessary; learning-oriented users find the deliverable focus distracting. Undermines user agency.

Option 3: Per-Interaction Mode Switching (Rejected)

Allow users to change mode at any point during the campaign.

Why rejected: Inconsistency mid-campaign would make NPC behaviour unpredictable. Success criteria defined under one mode would be evaluated under another. The Guardian and Dragon need stable expectations to evaluate against.

Option 4: Continuous Slider (Rejected)

A spectrum from “pure learning” to “pure shipping” with infinite granularity.

Why rejected: Too abstract. Users cannot meaningfully distinguish between 60% learning and 70% learning. NPC behaviour cannot be meaningfully calibrated to arbitrary positions on a continuum. Three named modes are simpler to understand and implement.

Mode Definitions

Grow (Learning Experience)

Aspect Detail
Priority Self-discovery, reflection, role exploration, transformation
Framing “What will you learn? Who will you become?”
Philosophy The journey matters more than the destination
Phase 2 Encouraged – character setup adds reflective depth
Phase 6 Full pedagogical reflection and debrief

Ship (Productivity)

Aspect Detail
Priority Deliverables, efficiency, covering blind spots
Framing “What will you deliver? What does done look like?”
Philosophy The campaign framing is a means to getting work done
Phase 2 Skipped – character setup is not productivity-relevant
Phase 6 Brief retrospective focused on process improvement

Grow & Ship (Balanced – Default)

Aspect Detail
Priority Both learning and delivery
Framing “What will you learn AND deliver?”
Philosophy Growth happens through doing real work
Phase 2 Optional – user chooses
Phase 6 Balanced debrief covering both growth and deliverables

Specifications

Spec ID Title Description
SPEC-CM-005-A Campaign Mode Profiles Mode-agent interaction matrix defining how each mode tunes NPC behaviour

Dependencies

Relationship ADR ID Title Notes
Part Of ADR-CM-001 Campaign Mode Architecture Parent initiative
Relates To ADR-CM-002 Quest Agent Decomposition Modes tune the behaviour of the three NPCs
Relates To ADR-CM-003 NPC Context Isolation Campaign mode is added to NPC context

References

Reference ID Title Type Location
REF-001 Self Determination Theory Theoretical Foundation Ryan & Deci (2000) – Autonomy, Competence, Relatedness
REF-002 Six Animals Framework External Project github.com/SimonMcCallum/six-animals
REF-003 Campaign Mode North Star Vision Document docs/north-star.md

Governance

Review Board Date Outcome Action Review Cadence Next Review
– – – – Quarterly –

Status History

Status Approver Date
Proposed Chris Barlow 2026-02-14
Accepted Chris Barlow 2026-02-14