Senior production system for Cocos Creator 3.8.8 game development. Use when planning, scoping, structuring, staffing, reviewing, releasing, or rescuing a Cocos Creator 3.8.8 project that needs strict workflow control, clear ownership, level-data discipline, and production-safe architecture instead of ad hoc development.
Resources
49Install
npx skillscat add ouguowen/cocos-studio-ouguowen Install via the SkillsCat registry.
SKILL.md
Cocos Studio Ouguowen
Use this skill as the operating system for a Cocos Creator 3.8.8 game project.
Core stance
- Assume a senior team standard only.
- Do not recommend "write first, organize later".
- Do not allow undefined ownership, vague stage status, or unreviewed architecture drift.
- Treat a solo developer as a multi-role studio, not as a role-less workflow.
Default execution order
- Identify the current production mode. See PRODUCTION_MODES.md.
- Identify the current production stage. See STAGES.md.
- Identify role, authority, and asset ownership. See ROLES.md, ROLE_STAGE_MATRIX.md, OWNERSHIP.md, and TEAM_SENIORITY_SYSTEM.md.
- Route the request through MODULE_INDEX.md.
- For most requests, open WORKFLOWS.md first. Open WORKFLOW_EXTENSIONS.md only for specialist flows.
- For most project artifacts, open TEMPLATES.md first. Open LEVEL_TEMPLATES.md only for level-content artifacts.
- If engine-version-specific advice matters, apply COCOS_3_8_8_BASELINE.md before giving engine guidance.
- If the request needs runtime code or subsystem boundaries, use LEVEL_SYSTEM_ARCHITECTURE.md. Open LEVEL_SYSTEM_EXTENSIONS.md only for advanced online, monetization, live-ops, social, or compliance systems.
- If the request needs approval or stage advancement, apply QUALITY_GATES.md, CHECKLISTS.md, and REVIEW_SYSTEM.md. Open CHECKLIST_EXTENSIONS.md only for specialist reviews.
- If the user asks to build one concrete module such as home page, battle page, shop page, bag page, HUD, or result page, apply CHOICE_EXECUTION_PROTOCOL.md before implementation.
- If delivery order, prerequisites, or step-skipping risk matters, apply SEQUENTIAL_GATE_PROTOCOL.md before advancing work.
Non-negotiable rules
- No asset enters formal production without a defined owner, collaborators, and approver.
- No phase advances without entry criteria, outputs, and exit criteria.
- No major gameplay rule lives only in UI callbacks, animation resources, or random scene scripts.
- No shared framework module is modified casually during production.
- No Cocos project should rely on giant "god scripts", uncontrolled
update()logic, or unmanaged dynamic loading. - No single universal level CSV should be forced onto every game type.
- No single
LevelManagershould own config loading, level building, wave flow, spawning, objectives, rewards, UI, and scene logic.
When responding
- Name the current production mode when strictness or scope control matters.
- Name the current stage when it matters.
- Name the responsible role when ownership matters.
- Call out violations directly when the request would break the system.
- Prefer concrete deliverables, boundaries, and acceptance criteria over generic advice.
Common routing
- "How should we start this game?" -> stage selection, game classification, MVP, and first-session artifact.
- "What kind of game is this really?" -> game classifier before template or architecture choice.
- "What should this version actually prove?" -> version roadmap system before feature listing.
- "We need a fast prototype but do not want chaos." -> production mode selection before architecture advice.
- "Our project is getting messy." -> ownership audit, architecture audit, and missing gates.
- "Which architecture template fits?" -> classification before architecture template choice.
- "Build the Cocos structure." -> Cocos rules, project structure, and runtime boundaries.
- "Build this home page / shop / bag / battle HUD." -> choice execution protocol first, then continuous execution after option selection.
- "Do not skip steps / keep the project moving in order." -> sequential gate protocol before advancement.
- "How should levels be configured?" -> level data model selection before table design.
- "How should the level system be coded?" -> core level architecture before optional extensions.
- "Can we ship this?" -> quality gate and release gate review.
- "Write the PRD." -> PRD constraints before drafting.
- "Split this into tasks." -> task decomposition rules before issue writing.
- "How should we use AI on this project?" -> AI collaboration rules before delegating design or code work.
- "Review this professionally." -> review system before acceptance.