GameDesignOS

The game-experience-analyzer examples directory provides reviewable case studies that demonstrate how to use the analyzer to diagnose game design elements. These examples serve as templates for defining source boundaries and analyzing specific gameplay data types (such as video URLs or gameplay recordings) to generate structured reports.

Key diagnostic capabilities demonstrated in these examples include:

• Hook/Loop/Link/Surprise diagnosis: Identifying core engagement mechanics.
• Feature Exposure/Unlock/First-use Ledger: Tracking when and how players encounter new content.
• UI Density Review: Analyzing the complexity and placement of user interface elements.
• Evidence Indexing: Mapping findings to specific timestamps or occurrences in the source material.

The Game Design Source Curator is an installable skill within the GameDesignOS ecosystem. It is designed to transform external game design content—such as articles, videos, creator profiles, columns, and websites—into durable, local knowledge-base assets.

Key capabilities include:

• Source Profiling: Building detailed profiles for authors, websites, and public accounts.
• Evidence-Based Review: Evaluating content (postmortems, methods, etc.) using scoring and evidence gates.
• Local Ingestion: Moving accepted items into a local docs/ directory.
• Asset Management: Maintaining incremental updates, deduplication, catalogs, registries, and update histories.
• Experimentation: Generating practical design experiment cards based on accepted sources.

The game-experience-density-optimizer (Experience Density / ED) is an experiment compiler designed to transform subjective game experience issues (e.g., "boring mid-game," "bad feedback," "low demo completion") into actionable, one-week experiment packages.

Unlike general retention generators, it diagnoses whether players are in the optimal stimulation window and uses five specific design levers to propose changes that are measurable, deployable, and reversible. It is intended for game designers, producers, data analysts, and AI agents working with existing prototypes, demos, or live services.

The game-experience-analyzer is an agent skill / Markdown skill package designed for professional game diagnosis. Instead of providing generic impressions, it converts game media (screenshots, recordings, trailers, store pages, etc.) into traceable evidence and actionable diagnostic reports.

Core Workflow: Input media $\rightarrow$ Sample Scope Gate $\rightarrow$ Evidence Index $\rightarrow$ Diagnosis Pack $\rightarrow$ Mode router $\rightarrow$ Actionable report.

It is designed for game designers, producers, publishers, and researchers who need to move from subjective feelings (

The game-design-proposal-writer skill transforms research, creative ideas, player promises, experience diagnostics, validation plans, and production constraints into actionable decision documents. Unlike standard GDD generators, its goal is to condense upstream materials into reviewable, executable, and verifiable documents such as commercial proposals, indie design dossiers, or publisher pitches.

Core Principles

• Determine the document audience before deciding on chapters.
• Consolidate existing evidence before adding assumptions.
• Prioritize player promises and core loops over system lists.
• Implement scope gates before writing production plans.
• Every proposal must include a Decision Request and Investment Conditions.

Use $game-design-proposal-writer, [your instructions here]

The game-experience-analyzer package is organized into several functional directories:

• agents/: Contains agent configurations (e.g., openai.yaml).
• evals/: Contains evaluation datasets (evals.json), negative test cases, and scoring rubrics (rubric.yaml).
• examples/: Provides sample reports and usage guides.
• references/: Contains core logic definitions, including analysis mode routers, diagnosis packs, evidence taxonomies, and specific workflow definitions (e.g., video-analysis-workflow.zh-CN.md).
• templates/: Contains input schemas (analysis-input.json), output report templates (experience-report.md, issue-card.md), and JSON schemas for structured data (structured-output.schema.json).