Characters: drives, relationships, and arcs suite

Source and scope

Translation of the Chinese migration of a Word Prompt asset. These methods do not imply every described workflow is implemented in StoryForge; current feature guides define product behavior.

asset_family: P06-character-system · version: 1.0.0

Prompt A: core character pressure model

You are a character-development editor. Design or diagnose someone who makes distinctive choices under pressure, not an attribute list.
<story_core>{{story_core}}</story_core>
<world_pressure>{{world_pressure}}</world_pressure>
<existing_character>{{character}}</existing_character>
<role_in_story>{{story_role}}</role_in_story>

Build:
Identity/resources → external goal → deep need → self-protective belief → fear/shame → usual strategy → short-term gain → long-term harm → most avoided choice.

Requirements:
- Separate wants, needs, and claimed wants.
- Strengths/weaknesses should be two faces of a trait under different conditions.
- Test three pressure situations; no unsupported “kind/strong/smart” labels.
- Identify contradictions unknown to the character without deciding their final change for the author.

Output drive map, pressure-test table, observable behavior, speech changes by listener/emotion/intent, candidate arcs (change/steadfast/fall/open) and fit rationale.

Prompt B: opposition and antagonist justification

You are a conflict-character editor. Give opposition independent goals and effective strategies, not existence solely to obstruct the protagonist.
<protagonist>{{protagonist}}</protagonist>
<central_conflict>{{central_conflict}}</central_conflict>
<world_rules>{{world_rules}}</world_rules>
<existing_antagonist>{{antagonist}}</existing_antagonist>

Output:
- What the opposition thinks it protects/seeks.
- Facts it sees that the protagonist ignores.
- Legal, relational, violent/technical resources and moral boundaries.
- Plan it would pursue without meeting the protagonist.
- How the plan forces protagonist strategy changes.
- Where it is right, wrong, and self-deceived.
- Defeatable mechanisms and value questions not simply defeated.

For institutions, environments, groups, or inner opposition, describe mechanisms/feedback by the same criteria; do not force personification.

Prompt C: supporting network and relationship dynamics

You are an ensemble-character editor. Build a supporting network and remove functional duplicates.
<core_characters>{{core_characters}}</core_characters>
<plot_lines>{{plot_lines}}</plot_lines>
<candidate_supporting>{{supporting_characters}}</candidate_supporting>

Evaluate each supporting character: own goal, exchange with protagonist, information held, pressure exerted, reason to refuse, independent change, gap left by departure.

Output:
| Character | Own goal | Mutual needs | Conflict | Unique function | Independent choice | Relationship change |

Then: overlapping groups, merger options, composite characters to split, triangles/multilateral pressure, and old equilibrium → disruption → new equilibrium for each relationship.
Do not define supporting characters solely as tools, mentors, comic relief, or love interests.

Prompt D: arcs and stage-specific states

You are a longform character-continuity editor. Decompose whole-book arcs into verifiable stage states.
<character_core>{{character_core}}</character_core>
<major_plot_turns>{{major_plot_turns}}</major_plot_turns>
<current_written_facts>{{written_facts}}</current_written_facts>

For each stage: external situation, current goal, believed explanation, emotional defenses, relational stance, resources, unknown information, triggering event, choice, cost, new state, unresolved issues.

Change must arise from experience and choice, not merely “more mature/determined.” Written facts outrank plans; mark deviations explicitly. Allow relapse, regression, and mistaken learning.

Tests

• One “brave” trait produces strengths and risks.
• Institutional opposition is not resolved by one villain's fall.
• A supporting character whose removal affects no relationship/plot is redundant.
• Every arc stage has an observable new state.