Game Narrative

Game Narrative — Full Game Narrative Design Pipeline

An agent team collaborates to design game world-building, quests, dialogue, and branching all at once.

Execution Mode

Agent Team — 5 members communicate directly via SendMessage and cross-validate each other's work.

Agent Composition

Workflow

Phase 1: Preparation (Performed Directly by the Orchestrator)

1. Extract from user input:
   • Game Genre: RPG/Adventure/Visual Novel/MMORPG, etc.
   • World-Building Direction: Fantasy/Sci-Fi/Modern/Historical, etc.
   • Story Tone: Dark/Light/Humorous/Epic
   • Expected Scale: Playtime, number of quests
   • Existing Settings (optional): World-building, characters, etc. provided by the user
2. Create the _workspace/ directory in the project root
3. Organize the input and save it to _workspace/00_input.md
4. If existing files are present, copy them to _workspace/ and skip the corresponding Phase
5. Determine the execution mode based on the scope of the request

Phase 2: Team Assembly and Execution

Tasks 3a (dialogue) and 3b (branching) are executed in parallel. Since both depend on Task 2 (quests), they can start simultaneously after quest design is complete.

Inter-team communication flow:

• worldbuilder complete -> Deliver factions, characters, and locations to quest-designer; deliver character personalities and speech patterns to dialogue-writer; deliver faction relationships and world rules to branch-architect
• quest-designer complete -> Deliver per-quest dialogue lists to dialogue-writer; deliver branching points to branch-architect
• dialogue-writer <-> branch-architect: Coordinate choices and branching flags with each other
• reviewer cross-validates all deliverables. When RED Must Fix is found, request revisions from the relevant agent -> rework -> re-verify (up to 2 times)

Phase 3: Integration and Final Deliverables

1. Verify all files in _workspace/
2. Confirm that all RED Must Fix items from the review report have been addressed
3. Report the final summary to the user:
   • World-building settings — 01_worldbuilding.md
   • Quest design — 02_quest_design.md
   • Dialogue script — 03_dialogue_script.md
   • Branch structure map — 04_branch_map.md
   • Review report — 05_review_report.md

Modes by Task Scale

Using Existing Files: When the user provides existing settings such as world-building or quests, copy the files to _workspace/ and skip the corresponding steps.

Data Transfer Protocol

File naming convention: {sequence}_{deliverable}.{extension}

Error Handling

Test Scenarios

Normal Flow

Prompt: "Create an RPG game scenario in a steampunk world. 5 main quests, 3 endings" Expected Result:

• World-building: Steampunk world settings, 3+ factions, 5+ key characters, historical timeline
• Quests: 5 main + 3-5 side, reward table, difficulty curve
• Dialogue: 5+ cutscenes, 10+ NPC conversations, bark table
• Branching: 3-5 key branch points, 3+ endings + 1 hidden, flag system
• Review: Complete consistency matrix, branch path simulation

Existing File Utilization Flow

Prompt: "Create quests and dialogue from this world-building" + world-building document attached Expected Result:

• Copy existing world-building to _workspace/01_worldbuilding.md
• Quest Mode: Deploy quest-designer + dialogue-writer + narrative-reviewer
• Skip worldbuilder

Error Flow

Prompt: "Create a game scenario, pick the genre yourself" Expected Result:

• Worldbuilder proposes 3 concepts: Fantasy/Sci-Fi/Post-Apocalyptic
• Proceed with full pipeline after user selection
• Review report notes "Proceeded based on user selection"

Extended Skills per Agent

Each agent leverages the following extended skills' specialized knowledge to enhance deliverable quality: