design red team audit
An adversarial design auditor that identifies failure modes and hidden risks in product pitches and game features.
$12
Works with the AI tools you already use
Claude Code
Cursor
Codex CLI
GitHub Copilot
Gemini CLI+20 moreClaude CodeCursorCodex CLIGitHub CopilotGemini CLI
OpenClaw
Manus
Grok+17 agentsdesign red team audit
Example session with this skill installed
Audit this feature: A 'Global Renown' system where players gain points for every action, appearing on a global leaderboard. We want to drive long-term retention and daily active users.
- Read your context and instructions
- Compiled the design red team
- Generated the UI component
I have audited the 'Global Renown' system. The concept is structurally weak as a retention driver.
Design Red Team Audit: Global Renown System
Verdict: Structurally weak
The system relies on a generic accumulation metric that lacks a specific gameplay loop or meaningful choice. It is likely to be perceived as 'background noise' by casual players and an exhausting treadmill by competitive players, leading to burnout rather than retention.
Most Credible Failure Modes
| Failure Mode | Why it happens | Likely Consequence | Early Warning Signs | Possible Mitigation |
|---|---|---|---|---|
| The 'Rich Get Richer' Wall | Early adopters or high-uptime players create an insurmountable lead in the first 72 hours. | New or returning players feel it is impossible to compete, ignoring the system entirely. | Leaderboard gap between top 1% and top 10% grows exponentially in week 1. | Implement seasonal resets or 'velocity' based scoring rather than lifetime totals. |
| Incentive Cannibalization | Players optimize for 'Renown-per-minute' instead of playing the fun parts of the game. | High-value, high-effort content is ignored in favor of repetitive, low-effort 'Renown farming.' | Player heatmaps show heavy clustering in low-level areas; engagement in diverse modes drops. | Weight Renown rewards based on activity difficulty and unique completion bonuses. |
| Feedback Opacity | Gaining points for 'every action' makes the reward feel meaningless and disconnected. | The system fails to provide a dopamine hit; players stop noticing the notifications. | Survey data shows players cannot explain how they earned their last 1,000 points. | Batch notifications; link Renown to specific, high-visibility 'Hero Actions' instead of everything. |
Weak Assumptions
Assumption: Players find global rankings inherently motivating without tiered rewards. (Status alone rarely drives long-term retention in non-hardcore segments).
*
Assumption: The 'Renown' economy won't be exploited by botting or macroing. (If every action counts, the most boring actions will be automated).
What Would Need To Be True
For this to succeed, the system must offer tangible power or utility that does not break the game balance, and it must include catch-up mechanisms that allow late-comers to feel competitive within their specific cohort.
Next steps
- Define 3 specific 'Hero Actions' that grant 80% of the Renown to prevent farming.
- Model the leaderboard spread for a player joining 3 months post-launch.
- Design a tiered 'Leagues' system to replace the single global list.
design-red-team-audit.tsx
TSX · React component
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
What you get
About this skill
The problem
Design pitches often suffer from confirmation bias and vague optimism. Teams spend months building features that look good on paper but collapse due to hidden systemic conflicts or misaligned player incentives.
What it does
- Performs an adversarial pre-mortem by assuming your feature or product idea has already failed.
- Identifies the 3-7 most credible failure mechanisms based on systemic design, UX comprehension, and production scalability.
- Exposes weak underlying assumptions that the team is likely taking for granted.
- Evaluates the MVP to determine if cutting 'connective tissue' will result in a false negative during testing.
- Proposes the fastest de-risking moves to validate uncertainties before significant resources are committed.
Why this beats prompting it yourself
Generic LLM prompts tend to be overly supportive or offer vague 'pros and cons' fluff. This skill is hard-coded to ignore flattery and hunt for specific failure mechanisms, providing a blunt, skeptical review that mimics an experienced lead designer or external auditor.
Use cases
- Stress-testing a game mechanic before moving from paper design to prototyping.
- Auditing a product roadmap to identify items that cannibalize existing systems.
- Performing a pre-mortem on a feature pitch to identify hidden production dependencies.
- Challenging a KPI strategy to ensure you aren't measuring activity while harming the user experience.
Known limitations
This skill requires a detailed feature description or pitch to be effective; vague inputs will result in broader, less actionable critiques.
How to install
Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.
- 1
Download the ZIP
Free skills download straight away. Paid skills unlock right after purchase.
- 2
Unzip into your skills folder
Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.
- 3
Ask your agent to use it
Restart the agent if it was already running. It picks the skill up automatically - no config needed.
Skills folder by agent
Click the path to copy it. Create the folder if it does not exist yet.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
30-day refund
Not a fit? Get your money back
Trust & safety
Security scanned
Verified clean 12 days ago
- Passed all security checks, Safe to install