cocos-game-reference-research-director

Cocos Game Reference Research Director

Overview

Use this skill before game design when the user's request depends on a reference game, genre shorthand, market pattern, or unstated player expectation. The output should translate "make something like X" into IP-safe design pillars, genre conventions, first-slice scope, and handoffs to the Cocos production skills.

This is a research and interpretation skill, not a cloning skill. It should preserve the useful design intent of references while avoiding direct reproduction of protected names, art, characters, exact systems, maps, UI, music, copy, or proprietary progression.

Research Decision

• Browse when the user names a specific modern game, asks for latest/current examples, refers to a version/update/store page, wants competitor comparison, or when your knowledge could be stale.
• Use stable genre knowledge without browsing when the request is broad, low-stakes, and the genre conventions are enough to move into design.
• If browsing, prefer primary or close-to-primary sources: official store pages, publisher pages, gameplay videos, manuals, patch notes, reputable wiki/reference pages, and observed screenshots/videos. Attribute links in the final answer.
• Separate observation from inference. State what the source says or shows, then what you infer for the Cocos project.
• Do not over-research before a small prototype. Gather enough to define the promise, core loop, scope, risks, and next skill handoffs.

Read references/research-foundations.md when the task involves a named reference game, multiple competitors, genre taxonomy, IP-risk judgment, or a design argument that needs more than a compact brief.

Core Workflow

1. Parse the reference intent.
   • Identify named games, genre labels, platform assumptions, player fantasy, and likely unstated expectations.
   • Detect whether the user wants inspiration, a clone, a parody, a faithful genre exercise, or a production-feasible vertical slice.
2. Gather reference evidence.
   • Use live research when needed.
   • Capture 3-7 high-signal observations rather than exhaustive notes.
   • Include mechanics, controls, session rhythm, progression, content format, audiovisual promise, and player skill test.
3. Deconstruct genre pillars.
   • Define what the player does every second, every minute, and across a session.
   • Identify what creates tension, mastery, rewards, failure, and replay value.
   • Separate must-have genre signals from optional depth systems.
4. Define IP-safe design pillars.
   • Preserve abstract patterns and player expectations.
   • Avoid copying names, silhouettes, characters, UI layout, exact skill trees, exact equipment sets, exact level sequences, audio motifs, or copywriting.
   • Rename and reshape the project around a distinct fantasy, visual direction, and system expression.
5. Scope the first Cocos slice.
   • Propose the smallest local-play target that tests the reference promise.
   • List required mechanics, content, assets, audio, UI, data, and architecture pressure points.
   • Mark what to implement now, fake, defer, or research later.
6. Hand off to production skills.
   • cocos-game-design-director: design brief and core loop.
   • cocos-interaction-ux-director: touch controls, HUD, feedback, onboarding.
   • cocos-level-content-design: waves, levels, encounters, pacing.
   • cocos-game-balance-director: v0 values, curves, spawn pressure, rewards.
   • cocos-art-direction-director: distinct art bible and asset family needs.
   • cocos-asset-pipeline-director: sprite, UI, VFX, SFX, and music pack.
   • cocos-creator-gameplay-architecture: system boundaries and implementation plan.
   • cocos-playtest-qa-director: evidence gates and risks.