Design System

是什么

帮你给某一个具体的系统（比如"会员等级"、"积分规则"、"消息中心"）写出一份完整、自洽、研发能直接动手的系统设计文档。它不是让你填表，而是一段一段地陪你把每个必备章节聊清楚，并自动把已有上下文（依赖系统、术语表）串进来。

怎么用

1. 先给它一个系统的短名字，让它去扫一遍现有文档，把相关上下文捞回来。
2. 跟着它一节一节往下推进：先讲清楚"玩家/用户的体感"，再谈规则、公式、边界、依赖。
3. 每过一节就让它落盘一次，避免聊到一半丢失进度。
4. 遇到拿不准的依赖系统，把对应文档丢给它，让它自动交叉引用，不要凭记忆写。
5. 最后让它跑一次完整自检，确认八大章节都齐了、术语统一了，再把文档给研发评审。

架构图

flowchart LR
  A[系统名] --> B[拉取相关上下文]
  B --> C[逐节协同撰写]
  C --> D[交叉引用依赖]
  D --> E[落盘系统设计文档]

When this skill is invoked:

1. Parse Arguments & Validate

A system name argument is required. If missing, fail with:

"Usage: /design-system <system-name> — e.g., /design-system combat-system Run /map-systems first to create the systems index, then use this skill to write individual system GDDs."

Normalize the system name to kebab-case for the filename (e.g., "combat system" becomes combat-system).

2. Gather Context (Read Phase)

Read all relevant context before asking the user anything. This is the skill's primary advantage over ad-hoc design — it arrives informed.

2a: Required Reads

• Game concept: Read design/gdd/game-concept.md — fail if missing:

  "No game concept found. Run /brainstorm first."

• Systems index: Read design/gdd/systems-index.md — fail if missing:

  "No systems index found. Run /map-systems first to map your systems."

• Target system: Find the system in the index. If not listed, warn:

  "[system-name] is not in the systems index. Would you like to add it, or design it as an off-index system?"

2b: Dependency Reads

From the systems index, identify:

• Upstream dependencies: Systems this one depends on. Read their GDDs if they exist (these contain decisions this system must respect).
• Downstream dependents: Systems that depend on this one. Read their GDDs if they exist (these contain expectations this system must satisfy).

For each dependency GDD that exists, extract and hold in context:

• Key interfaces (what data flows between the systems)
• Formulas that reference this system's outputs
• Edge cases that assume this system's behavior
• Tuning knobs that feed into this system

2c: Optional Reads

• Game pillars: Read design/gdd/game-pillars.md if it exists
• Existing GDD: Read design/gdd/[system-name].md if it exists (resume, don't restart from scratch)
• Related GDDs: Glob design/gdd/*.md and read any that are thematically related (e.g., if designing "status-effects", also read "combat-system" even if it's not a direct dependency)

2d: Present Context Summary

Before starting design work, present a brief summary to the user:

Designing: [System Name]

• Priority: [from index] | Layer: [from index]
• Depends on: [list, noting which have GDDs vs. undesigned]
• Depended on by: [list, noting which have GDDs vs. undesigned]
• Existing decisions to respect: [key constraints from dependency GDDs]
• Pillar alignment: [which pillar(s) this system primarily serves]

If any upstream dependencies are undesigned, warn:

"[dependency] doesn't have a GDD yet. We'll need to make assumptions about its interface. Consider designing it first, or we can define the expected contract and flag it as provisional."

Use AskUserQuestion:

• "Ready to start designing [system-name]?"
  • Options: "Yes, let's go", "Show me more context first", "Design a dependency first"

3. Create File Skeleton

Once the user confirms, immediately create the GDD file with empty section headers. This ensures incremental writes have a target.

Use the template structure from .claude/docs/templates/game-design-document.md:

# [System Name]

> **Status**: In Design
> **Author**: [user + agents]
> **Last Updated**: [today's date]
> **Implements Pillar**: [from context]

## Overview

[To be designed]

## Player Fantasy

[To be designed]

## Detailed Design

### Core Rules

[To be designed]

### States and Transitions

[To be designed]

### Interactions with Other Systems

[To be designed]

## Formulas

[To be designed]

## Edge Cases

[To be designed]

## Dependencies

[To be designed]

## Tuning Knobs

[To be designed]

## Visual/Audio Requirements

[To be designed]

## UI Requirements

[To be designed]

## Acceptance Criteria

[To be designed]

## Open Questions

[To be designed]

Ask: "May I create the skeleton file at design/gdd/[system-name].md?"

After writing, update production/session-state/active.md with:

• Task: Designing [system-name] GDD
• Current section: Starting (skeleton created)
• File: design/gdd/[system-name].md

4. Section-by-Section Design

Walk through each section in order. For each section, follow this cycle:

The Section Cycle

Context  ->  Questions  ->  Options  ->  Decision  ->  Draft  ->  Approval  ->  Write

1. Context: State what this section needs to contain, and surface any relevant decisions from dependency GDDs that constrain it.

2. Questions: Ask clarifying questions specific to this section. Use AskUserQuestion for constrained questions, conversational text for open-ended exploration.

3. Options: Where the section involves design choices (not just documentation), present 2-4 approaches with pros/cons. Explain reasoning in conversation text, then use AskUserQuestion to capture the decision.

4. Decision: User picks an approach or provides custom direction.

5. Draft: Write the section content in conversation text for review. Flag any provisional assumptions about undesigned dependencies.

6. Approval: Ask "Approve this section, or would you like changes?"

7. Write: Use the Edit tool to replace the [To be designed] placeholder with the approved content. Confirm the write.

After writing each section, update production/session-state/active.md with the completed section name.

Section-Specific Guidance

Each section has unique design considerations and may benefit from specialist agents:

Section A: Overview

Goal: One paragraph a stranger could read and understand.

Questions to ask:

• What is this system in one sentence?
• How does a player interact with it? (active/passive/automatic)
• Why does this system exist — what would the game lose without it?

Cross-reference: Check that the description aligns with how the systems index describes it. Flag discrepancies.

Section B: Player Fantasy

Goal: The emotional target — what the player should feel.

Questions to ask:

• What emotion or power fantasy does this serve?
• What reference games nail this feeling? What specifically creates it?
• Is this a "system you love engaging with" or "infrastructure you don't notice"?

Cross-reference: Must align with the game pillars. If the system serves a pillar, quote the relevant pillar text.

Section C: Detailed Design (Core Rules, States, Interactions)

Goal: Unambiguous specification a programmer could implement without questions.

This is usually the largest section. Break it into sub-sections:

1. Core Rules: The fundamental mechanics. Use numbered rules for sequential processes, bullets for properties.
2. States and Transitions: If the system has states, map every state and every valid transition. Use a table.
3. Interactions with Other Systems: For each dependency (upstream and downstream), specify what data flows in, what flows out, and who owns the interface.

Questions to ask:

• Walk me through a typical use of this system, step by step
• What are the decision points the player faces?
• What can the player NOT do? (Constraints are as important as capabilities)

Agent delegation: For complex mechanics, use the Task tool to delegate to game-designer for high-level design review, or systems-designer for detailed mechanical modeling. Provide the full context gathered in Phase 2.

Cross-reference: For each interaction listed, verify it matches what the dependency GDD specifies. If the dependency says "damage is calculated as X" and this system expects something different, flag the conflict.

Section D: Formulas

Goal: Every mathematical formula, with variables defined, ranges specified, and edge cases noted.

Questions to ask:

• What are the core calculations this system performs?
• Should scaling be linear, logarithmic, or stepped?
• What should the output ranges be at early/mid/late game?

Agent delegation: For formula-heavy systems (combat, economy, progression), delegate to systems-designer via the Task tool. Provide:

• The Core Rules from Section C (already written to file)
• Tuning goals from the user
• Balance context from dependency GDDs

The agent should return proposed formulas with variable tables and expected output ranges. Present these to the user for review before approving.

Cross-reference: If a dependency GDD defines a formula whose output feeds into this system, reference it explicitly. Don't reinvent — connect.

Section E: Edge Cases

Goal: Explicitly handle unusual situations so they don't become bugs.

Questions to ask:

• What happens at zero? At maximum? At negative values?
• What happens when two effects trigger simultaneously?
• What happens if the player tries to exploit this? (Identify degenerate strategies)

Agent delegation: For systems with complex interactions, delegate to systems-designer to identify edge cases from the formula space. For narrative systems, consult narrative-director for story-breaking edge cases.

Cross-reference: Check edge cases against dependency GDDs. If combat says "damage cannot go below 1" but this system can reduce damage to 0, that's a conflict to resolve.

Section F: Dependencies

Goal: Map every system connection with direction and nature.

This section is partially pre-filled from the context gathering phase. Present the known dependencies from the systems index and ask:

• Are there dependencies I'm missing?
• For each dependency, what's the specific data interface?
• Which dependencies are hard (system cannot function without it) vs. soft (enhanced by it but works without it)?

Cross-reference: This section must be bidirectionally consistent. If this system lists "depends on Combat", then the Combat GDD should list "depended on by [this system]". Flag any one-directional dependencies for correction.

Section G: Tuning Knobs

Goal: Every designer-adjustable value, with safe ranges and extreme behaviors.

Questions to ask:

• What values should designers be able to tweak without code changes?
• For each knob, what breaks if it's set too high? Too low?
• Which knobs interact with each other? (Changing A makes B irrelevant)

Agent delegation: If formulas are complex, delegate to systems-designer to derive tuning knobs from the formula variables.

Cross-reference: If a dependency GDD lists tuning knobs that affect this system, reference them here. Don't create duplicate knobs — point to the source of truth.

Section H: Acceptance Criteria

Goal: Testable conditions that prove the system works as designed.

Questions to ask:

• What's the minimum set of tests that prove this works?
• What performance budget does this system get? (frame time, memory)
• What would a QA tester check first?

Cross-reference: Include criteria that verify cross-system interactions work, not just this system in isolation.

Optional Sections: Visual/Audio, UI Requirements, Open Questions

These sections are included in the template but aren't part of the 8 required sections. Offer them after the required sections are done:

Use AskUserQuestion:

• "The 8 required sections are complete. Do you want to also define Visual/Audio requirements, UI requirements, or capture open questions?"
  • Options: "Yes, all three", "Just open questions", "Skip — I'll add these later"

For Visual/Audio: Coordinate with art-director and audio-director if detail is needed. Often a brief note suffices at the GDD stage.

For UI Requirements: Coordinate with ux-designer for complex UI systems.

For Open Questions: Capture anything that came up during design that wasn't fully resolved. Each question should have an owner and target resolution date.

5. Post-Design Validation

After all sections are written:

5a: Self-Check

Read back the complete GDD from file (not from conversation memory — the file is the source of truth). Verify:

• All 8 required sections have real content (not placeholders)
• Formulas reference defined variables
• Edge cases have resolutions
• Dependencies are listed with interfaces
• Acceptance criteria are testable

5b: Offer Design Review

Present a completion summary:

GDD Complete: [System Name]

• Sections written: [list]
• Provisional assumptions: [list any assumptions about undesigned dependencies]
• Cross-system conflicts found: [list or "none"]

Use AskUserQuestion:

• "Run /design-review now to validate the GDD?"
  • Options: "Yes, run review now", "I'll review it myself first", "Skip review"

If yes, invoke the design-review skill on the completed file.

5c: Update Systems Index

After the GDD is complete (and optionally reviewed):

• Read the systems index
• Update the target system's row:
  • If design-review was run and verdict is APPROVED: Status → "Approved"
  • If design-review was run and verdict is NEEDS REVISION: Status → "In Review"
  • If design-review was skipped: Status → "Designed" (pending review)
  • If the user chose "I'll review it myself first": Status → "Designed"
  • Design Doc: link to design/gdd/[system-name].md
• Update the Progress Tracker counts

Ask: "May I update the systems index at design/gdd/systems-index.md?"

5d: Update Session State

Update production/session-state/active.md with:

• Task: [system-name] GDD
• Status: Complete (or In Review if design-review was run)
• File: design/gdd/[system-name].md
• Sections: All 8 written
• Next: [suggest next system from design order]

5e: Suggest Next Steps

Use AskUserQuestion:

• "What's next?"
  • Options:
    • "Design next system ([next-in-order])" — if undesigned systems remain
    • "Fix review findings" — if design-review flagged issues
    • "Stop here for this session"
    • "Run /gate-check" — if enough MVP systems are designed

6. Specialist Agent Routing

This skill delegates to specialist agents for domain expertise. The main session orchestrates the overall flow; agents provide expert content.

When delegating via Task tool:

• Provide: system name, game concept summary, dependency GDD excerpts, the specific section being worked on, and what question needs expert input
• The agent returns analysis/proposals to the main session
• The main session presents the agent's output to the user via AskUserQuestion
• The user decides; the main session writes to file
• Agents do NOT write to files directly — the main session owns all file writes

7. Recovery & Resume

If the session is interrupted (compaction, crash, new session):

1. Read production/session-state/active.md — it records the current system and which sections are complete
2. Read design/gdd/[system-name].md — sections with real content are done; sections with [To be designed] still need work
3. Resume from the next incomplete section — no need to re-discuss completed ones

This is why incremental writing matters: every approved section survives any disruption.

Collaborative Protocol

This skill follows the collaborative design principle at every step:

1. Question -> Options -> Decision -> Draft -> Approval for every section
2. AskUserQuestion at every decision point (Explain -> Capture pattern):
   • Phase 2: "Ready to start, or need more context?"
   • Phase 3: "May I create the skeleton?"
   • Phase 4 (each section): Design questions, approach options, draft approval
   • Phase 5: "Run design review? Update systems index? What's next?"
3. "May I write to [filepath]?" before the skeleton and before each section write
4. Incremental writing: Each section is written to file immediately after approval
5. Session state updates: After every section write
6. Cross-referencing: Every section checks existing GDDs for conflicts
7. Specialist routing: Complex sections get expert agent input, presented to the user for decision — never written silently

Never auto-generate the full GDD and present it as a fait accompli. Never write a section without user approval. Never contradict an existing approved GDD without flagging the conflict. Always show where decisions come from (dependency GDDs, pillars, user choices).