# Context

I'm building **[project name]** — [brief description of what you are building and the core experience you want players/users to have].

**Stack:** [your tech stack — languages, frameworks, rendering pipeline, deployment target]

**What already exists:**

- [list systems, shaders, mechanics, or features you have already implemented]

- [be specific — names of passes, uniforms, modules, or subsystems help]

**What's wrong:** [describe the gap between what exists today and the feel or quality you are aiming for. What reads as superficial or broken despite the underlying work?]

# What I want from you

Act as a senior [keyword] engineer specialising in [specific sub-domain — e.g. real-time graphics, game feel, distributed systems].

## Part 1 — Diagnosis framework

Explain what separates best-in-class implementations in this domain from surface-level or "themed" approximations. Cover the key dimensions relevant to [keyword], such as:

- **Core technical pillars** (the physics, optics, algorithms, or interaction models that define authenticity)

- **Transitions & boundaries** (edge cases, state changes, threshold effects)

- **Feel & responsiveness** (movement, feedback, pacing, camera, input)

- **Feedback loops** (audio, HUD, haptics, or UI that reinforces the core experience)

- **Life & scale** (ambient motion, entity behaviour, environmental detail, sense of place)

For each dimension, give a 1–10 "impact on feel" score specifically for [keyword].

## Part 2 — Open-source resources

List open-source repos and reference implementations I can learn from or port techniques from. Organise into priority tiers based on my stack:

1. **[Primary stack — highest priority, directly portable]**

2. **[Secondary engine/framework references — shader, post-process, or architecture patterns]**

3. **[Alternative engines — technique reference even if not directly portable]**

4. **[Other readable implementations — easy to port GLSL, algorithms, or data structures]**

5. **Papers & articles** (canonical write-ups, GPU Gems chapters, blog posts from respected practitioners)

For each entry include:

- Repo/link

- Licence

- What it does well

- **Specific technique to extract** (e.g. "Beer–Lambert fog in fragment shader using view depth")

- **Port difficulty:** Easy / Medium / Hard for my stack

- **Performance cost** on my target platform (low/medium/high)

## Part 3 — Prioritised action plan for my codebase

Given my existing systems (listed above), recommend a **top 10 improvement list** ordered by impact/effort. For each item:

- What's missing today

- Which reference repo implements it best

- Where it would likely plug into my codebase (post-process pass, material uniform, physics tick, audio bus, etc.)

- Rough implementation sketch (pseudocode or shader snippet outline, not a full rewrite)

## Part 4 — Anti-patterns

List common mistakes in this domain on my target platform (e.g. uniform colour washes, missing channel falloff, static decals pretending to be dynamic effects, entities that move like drones). For each, explain how to avoid it.

# Constraints

- Must stay in [your stack — no migration to other engines unless noted as reference-only]

- Target [performance profile — e.g. mid-range laptops, mobile browsers]; need scalable quality tiers

- Aesthetic/environment constraints: [e.g. NZ-themed reef, enterprise SaaS minimalism, retro pixel art]

- Files should stay modular (~500 lines); prefer extending existing systems over monolithic rewrites

# Output format

Use tables where helpful. Be specific and actionable — I want a roadmap I can implement, not generic "add more polish" advice.